Seminář z umělé inteligence I a II
historie semináře

Roman Barták, KTIML

This is a page showing history of the seminar. The content of the current seminar is available here.
Seminar    LS 2023/2024 (NAIL052):
Úterý (Tuesday) 14:00 - 15:30, lecture room S6 (Malá Strana, 2nd floor)

Seminar will run in Czech or English (depending on students attending) and it will be organized as a series of presentations of research papers followed by Q&A and evaluation of each presentation OR solving some Kaggle challenge.

For presentations, we will focus on hierarchical (HTN) planning. We expect to cover the following topics:

  • semantics of hierarchical planning models
  • hierarchical planning
  • hierarchical plan verification
  • hierarchical plan correction
  • hierarchical plan recognition
  • hierarchical plan repair
  • learning HTN planning models

This is a list of papes to start with; students can also (and are expected) bring own papers, but they must be approved by the teacher:

  1. Kristýna Pantucková, Roman Barták: Using Earley Parser for Recognizing Totally Ordered Hierarchical Plans. ECAI 2023: 1819-1826
  2. Songtuan Lin, Gregor Behnke, Simona Ondrcková, Roman Barták, Pascal Bercher: On Total-Order HTN Plan Verification with Method Preconditions - An Extension of the CYK Parsing Algorithm. AAAI 2023: 12041-12048
  3. Simona Ondrcková, Roman Barták: On Semantics of Hierarchical Planning Domain Models with Decomposition Constraints and Empty Methods. ICTAI 2023: 349-353
  4. Roman Barták, Simona Ondrcková, Gregor Behnke, Pascal Bercher: On the Verification of Totally-Ordered HTN Plans. ICTAI 2021: 263-267
  5. Roman Barták, Simona Ondrcková, Gregor Behnke, Pascal Bercher: Correcting Hierarchical Plans by Action Deletion. KR 2021: 99-109
  6. Roman Barták, Simona Ondrcková, Adrien Maillard, Gregor Behnke, Pascal Bercher: A Novel Parsing-based Approach for Verification of Hierarchical Plans. ICTAI 2020: 118-125
  7. Roman Barták, Adrien Maillard, Rafael Caue Cardoso: Validation of Hierarchical Plans via Parsing of Attribute Grammars. ICAPS 2018: 11-19
  8. Chad Hogg, Hector Munoz-Avila, and Ugur Kuter: HTN-Maker: Learning HTNs with Minimal Additional Knowledge Engineering Required. In Proceedings of the Twenty-Third AAAI Conference on Artificial Intelligence (AAAI-08). AAAI Press.
  9. D. Nau, T.-C. Au, O. Ilghami, U. Kuter, W. Murdock, D. Wu, and F.Yaman: SHOP2: An HTN Planning System. JAIR, volume 20, pp. 379-404, 2003.
  10. Robert P. Goldman, Ugur Kuter, and Richard G. Freedman. Stable plan repair for state-space HTN planning. HPlan Workshop 2020

This is a preliminary program of the seminar.
20.02. 2024 Kick-off meeting (starting at 14:30!!)  
27.02. 2024 Distribution of papers/talks (starting at 14:30!!)  
05.03. 2024 AAAI/IAAI Invited Talk: Objective-Driven AI: Towards Machines that can Learn, Reason, and Plan by Yann LeCun [working paper]
12.03. 2024

Challenge introduction (V.P)

19.03. 2024 Challenge introduction (M.C.)
26.03. 2024 Introduction to Multi-Agent Path Finding MAPF tutorial
02.04. 2024

MAPF Solving approaches

MAPF tutorial
09.04. 2024 Midterm report (V.P.; M.C.)  
16.04. 2024 Fireside Chat with Daniel Kahneman @ AAAI 2019 [YouTube]  
23.04. 2024

TBA

30.04. 2024 TBA
  
07.05. 2024 Final report and solution (V.P.; M.C.)  
14.05. 2024 cancelled (Rector Sports Day)
21.05. 2024 TBA  
Seminar     ZS 2023/2024 (NAIL004):
Úterý (Tuesday) 09:00 - 10:30, lecture room S6 (Malá Strana, 2nd floor)

Seminar will run in English (Czech on demand) and it will be organized as a series of Oxford-style debates on various AI-related topics (the topics discussed in recent years can be found at the seminar history page). Each side of the debate will be represented by two students (two "for" and two "against") pluse there will be a moderator. The side to which a student is allocated does not necessarily reflect his/her personal opinion!

Debate schedule:

  • prior debate:
    • Each side is requested to send one reference (web link, paper, etc.) to the teacher at least one week before the debate and this reference will be publicly available through the seminar web page.
  • debate day (Tuesday):
    • introduction of the topic by the moderator (5-10 minutes)
    • initial anonymous voting of audience (the result will be revealed after the discussion)
    • opening remarks (each speaker will have 5 minutes for the initial statement supporting his/her side; the order of sides is selected randomly at the beginning, speakers from both sides speak on a rota basis)
    • intra-panel discussion (between the speakers and the moderator with chance to react to the other side; 10-20 minutes)
    • Q&A (questions/comments from the audience with response from the speakers;  20-30 minutes)
    • closing remarks (each speaker will have 2 minutes; the order is reverse to the opening order)
    • final anonymous voting of audience
    • decision of the winner (the side with the increase of number of votes wins)

Between the Oxford Debates we will include Hackathons solving selected technical problems.

To get the credit, student is supposed to participate in at least one debate and to actively participate in at least eight seminars.

  
03.10. 2023 Cancelled!!  
10.10. 2023 Kick-off meeting   
17.10. 2023 Groups formation and topics distribution
24.10. 2023

Cancelled!
31.10. 2023 Oxford Debate
Should AI be used in the visa issuing process?
[Public procurement] [Public sector] [Public sector Review] [Public governance] [Public administration] [Bureaucracy] [Bureaucracy 2] [Bureaucratic transformation] 		M.M., R.P. (for); P.Y., K.K. (against)
before: 87% (for) - 13% (against)
after: 60% (for) - 40% (against)
07.11. 2023 Cancelled!  
14.11. 2023

Oxford Debate
If autonomous cars are proven to be statistically safer (and are used by the majority), should we outlaw humans from driving in regular traffic?
[Autonomy Levels] [Human vs. Autonomy] [Dangers] [EU Report] [Safety Overview] [Safety Comparison] [Future]

E.C., N.T. (for); A.M., S.P. (against)
before: 33% (for) - 67% (against)
after: 70% (for) - 30% (against)
21.11. 2023 Hackathon
How can AI help in personal travel (business/vacation)?
  • Team A: AI travel assistant with real-time updates
  • Team B: Planning a good trip completely
  • Team C: AIdea: Webapp for trip planning
  • Team D: Fully virtual tourism
    		•
    28.11. 2023 Oxford Debate
    Should research and development of general models more capable than GPT4 be paused?
    [AI regulation?] [AI regulation in EU] [AI pause] [AI open network] [Pause is mistake] [AI needed for Climate Crisis]     		F.B., V.B. (for); P.D., J.P. (againts)
    before: 22% (for) - 78% (against)
    after: 14% (for) - 86% (against)
    05.12. 2023 Hackathon
    How can we improve trustworthiness of AI systems?
  • Team A: user confidence; market evaluation
  • Team B: user education and market
  • Team C: understanding limitations, measuring correctness
  • Team D: comparison to human, certificates, transparency
    		•
    12.12. 2023 Oxford Debate
    Will AI techniques destroy more job positions than they create?
    [OECD Report] [Industry Perspective] [Future of Work] [Comparative Analysis] [Jobs Affected by AI] [BBC Report 1] [BBC Report 2] [WEF Study]     		J.Č., J.V. (for); A.A., P.Y. (against)
    before: 67% (for) - 33% (against)
    after: 55% (for) - 45% (against)
    19.12. 2023 Hackathon
    How can AI improve live in cities?
  • Team A: planning a new city from scratch
  • Team B: new concept of public transport (with different levels)
  • Team C: improving current public transport by better reaction to up-to-date situation
    		•   
    09.01. 2024 Oxford Debate
    Should AI entitites have rights and be protected?
    [The Hard Problem of Consciousness] [Consciousness in AI] [Recognizing Consciousness] [AI and Law] [AI and Human Rights]     		A.A., P.Y. (for); K.E., J.N. (against)
    before: 36% (for) - 64% (against)
    after: 27% (for) - 73% (against
    Seminar    LS 2022/2023 (NAIL052):
    Úterý (Tuesday) 09:00 - 10:30, lecture room S6 (Malá Strana, 2nd floor)

    Seminar will run in English and it will be organized as a series of presentations of research papers followed by Q&A and evaluation of each presentation.

    We will focus on large language models and their usage in general (beyond language), specifically on ChatGPT and similar techniques. This is a list of papes to select from; students can also bring own papers, but they must be approved by the teacher:

    1. Attention Is All You Need [paper]
    2. Language Models are Few-Shot Learners [paper]
    3. A Survey of Transformers [paper]
    4. Efficient Transformers: A Survey [paper]
    5. A Generalist Agent [paper]
    6. An Analysis of the Automatic Bug Fixing Performance of ChatGPT [paper]
    7. Toolformer: Language Models Can Teach Themselves to Use Tools [paper]
    8. How Good Are GPT Models at Machine Translation? A Comprehensive Evaluation [paper]
    9. Attention-based C-BiLSTM for fake news detection [paper]
    10. A benchmark study of machine learning models for online fake news detection [paper]
    11. Dissociating language and thought in large language models: a cognitive perspective [paper]
    12. Sparks of Artificial General Intelligence: Early experiments with GPT-4 [paper]

    HackAPrompt - a challenge to trick large language models worth money prize for winners.
    14.02. 2023 Kick-off meeting  
    21.02. 2023 Distribution of papers/talks   
    28.02. 2023 ChatGPT experiments (each student will prepare two examples, one showing how the system can solve a non-trivial problem and one showing how the system makes a significant mistake)             
    07.03. 2023

    ChatGPT experiments (each student will generate a short research report using ChatGPT and analyse its quality)
    14.03. 2023 Maroš Bratko: An Analysis of the Automatic Bug Fixing Performance of ChatGPT (6)
    21.03. 2023 each student will prepare three examples of ChatGPT solving a real short homework: 1) high school level 2) college level formally specified assignment 3) college level wordy assignment  
    28.03. 2023

    Michal Pospěch: Toolformer: Language Models Can Teach Themselves to Use Tools (7)

    04.04. 2023 cancelled
    11.04. 2023 Yuliia Alpaieva: How Good Are GPT Models at Machine Translation? A Comprehensive Evaluation (8)
    18.04. 2023

    Discussion on "Pause Giant AI Experiments: An Open Letter" (see at web)
    Survival Strategies for Depressed AI Academics (paper)

    25.04. 2023 Diellor Hoxhaj: Sparks of Artificial General Intelligence: Early experiments with GPT-4 (12), short
    		  
    02.05. 2023 Rem Lohinov: Attention Is All You Need (1)
    09.05. 2023 Marianna Malaireu: Fake news detection by using language modes (9,10)
    16.05. 2023 Diellor Hoxhaj: Sparks of Artificial General Intelligence: Early experiments with GPT-4 (12), full
    Seminar   ZS 2022/2023 (NAIL004):
    Úterý (Tuesday) 9:00 - 10:30, lecture room S8 (Malá Strana, 1st floor)

    Seminar will run in English (Czech on demand) and it will be organized as a series of Oxford-style debates on various AI-related topics (the topics discussed in recent years can be found at the seminar history page). Each side of the debate will be represented by two students (two "for" and two "against") pluse there will be a moderator. The side to which a student is allocated does not necessarily reflect his/her personal opinion!

    Debate schedule:

    • prior debate:
      • Each side is requested to send one reference (web link, paper, etc.) to the teacher at least one week before the debate and this reference will be publicly available through the seminar web page.
    • debate day (Tuesday):
      • introduction of the topic by the moderator (5-10 minutes)
      • initial anonymous voting of audience (the result will be revealed after the discussion)
      • opening remarks (each speaker will have 5 minutes for the initial statement supporting his/her side; the order of sides is selected randomly at the beginning, speakers from both sides speak on a rota basis)
      • intra-panel discussion (between the speakers and the moderator with chance to react to the other side; 10-20 minutes)
      • Q&A (questions/comments from the audience with response from the speakers;  20-30 minutes)
      • closing remarks (each speaker will have 2 minutes; the order is reverse to the opening order)
      • final anonymous voting of audience
      • decision of the winner (the side with the increase of number of votes wins)

    Between the Oxford Debates we will include Hackathons solving selected technical problems.
    04.10. 2022 canceled  
    11.10. 2022 Kick-off meeting  
    18.10. 2022 Intro to AI research at KTIML, team formation 
    25.10. 2022

    Oxford debate
    Should automated trading systems be allowed to trade on financial markets?
    [Algorithmic Trading] [High Frequency Trading] [HFT and stability] [HFT and Crashes] [Flash Crash 2010] [Crash 2010 video]

    		 P.M., M.S. (for); K.B.O., M.Ch. (against)
    before: 78% (for) - 22% (against)
    after: 82% (for) - 18% (against)
    01.11. 2022

    Hackathon
    How can AI help with efficent energy usage?

    • Team A: optimizing industrial heating system
    • Team B: adaptive street lights optimization
    • Team C: balancing energy production and consumption
    • Team D: smart house
    		  
    08.11. 2022 Oxford debate
    Should AI be allowed to kill?
    [Lethal Autonomous Weapons] [AI Arms Race] [Future of Warfare with AI] [Slaughterbots video]    		 R.G., T.T (for); S.R., M.Z. (against)
    before: 37% (for) - 63% (against)
    after: 25% (for) - 75% (against)
    15.11. 2022

    Hackathon
    How to eliminate AI Bias?

    • Team A: bias in computer vision (identifying examples out of domain)
    • Team B: bias in ML (human in a loop, education of users)
    • Team C: bias in medical field (better collection of data)
    		  
    22.11. 2022 Oxford debate
    Should AI be used for monitoring employee performance at work?
    [Monitoring Remote Workers] [Workforce Surveillance]     		Y.A., M.M. (for); G.K., M.P. (against)
    before: 38% (for) - 62% (against)
    after: 33% (for) - 67% (against)
    29.11. 2022 Hackathon
    How can AI help students?
    • Team A: a search tool for study materials
    • Team B: constructing individual study plans
    • Team C: automated student tutor
    • Team D: extracting lecture notes from videolectures
    06.12. 2022

    Oxford debate
    Should user own AI-generated art?
    [DeviantArt] [AI in Art Contest] [AI Copyright] [AI Art in Copyright Law]

    		L.A., E.A. (for); R.T., D.W (against)
    before: 50% (for) - 50% (against)
    after: 64% (for) - 36% (against)
    13.12. 2022 Hackathon
    How can AI improve life in cities?
    • Team A: smart toll
    • Team B: automated garbarge collection
    • Team C: automated individual public transport
    		  
    20.12. 2022 Oxford debate
    Should fully autonomous cars be allowed?
    [Pros&Const] [Benefits] [Ban] [Technology]     		U.O.D. (for); P.M. (against)
    before: 60% (for) - 40% (against)
    after: 70% (for) - 40% (against)
    03.01. 2023 canceled  
    Seminar    LS 2021/2022 (NAIL052):
    Středa (Wednesday) 12:20 - 13:50, lecture room S10 (Malá Strana, 1st floor)

    Seminar will run in Czech (English on demand) and it will be organized as a series of presentations of research papers followed by Q&A and evaluation of each presentation. There is also a classroom in Microsoft Teams for asynchronous discusions.

    We will focus on coordination of multiple agents, such as multi-agent pathfinding. This is a list of papes to select from; students can also bring own papers, but they must be approved by the teacher:

    1. Multi-Agent Pathfinding: Definitions, Variants, and Benchmark [paper]
    2. Finding Optimal Solutions to Cooperative Pathfinding Problems [paper]
    3. Conflict-Based Search For Optimal Multi-Agent Path Finding [paper]
    4. On SAT-Based Approaches for Multi-Agent Path Finding with the Sum-of-Costs Objective [paper]
    5. Scalable Rail Planning and Replanning: Winning the 2020 Flatland Challenge [paper]
    6. Lifelong Multi-Agent Path Finding in Large-Scale Warehouses [paper]
    7. Autonomous and Semi-Autonomous Intersection Management: A Survey [paper]
    8. Multi-Agent Pathfinding with Real-Time Heuristic Search [paper]
    9. PRIMAL: Pathfinding via Reinforcement and Imitation Multi-Agent Learning [paper]
    10. Branch-and-Cut-and-Price for Multi-Agent Pathfinding [paper]
    11. Cooperative Multi-Robot Navigation in Dynamic Environment with Deep Reinforcement Learning [paper]
    12. Trajectory Planning for Quadrotor Swarms [paper]
    13. Decentralized Path Planning for Multi-Agent Teams in Complex Environments using Rapidly-exploring Random Trees [paper]
    14. Efficient path planning for UAV formation via comprehensively improved particle swarm [paper]
    15. Robust Multi-Agent Path Finding and Executing [paper]
    16. Online Multi-Agent Pathfinding [paper]
    17. Multi-agent path finding on real robots [paper]
    18. A Scheduling-Based Approach to Multi-Agent Path Finding with Weighted and Capacitated Arcs [paper]
    19. Safe Multi-Agent Pathfinding with Time Uncertainty [paper]
    16.02. 2022 Kick-off meeting  
    23.02. 2022 Introduction to Multi-Agent Pathfinding  
    02.03. 2022 Jiří Krejčí: Conflict-Based Search For Optimal Multi-Agent Path Finding (3)
    09.03. 2022

    Radovan Tomala: Efficient path planning for UAV formation via comprehensively improved particle swarm (14)
    16.03. 2022 Gabriela Suchopárová: Cooperative Multi-Robot Navigation in Dynamic Environment with Deep Reinforcement Learning (11)                   
    23.03. 2022 canceled  
    30.03. 2022

    Renáta Pivodová: PRIMAL: Pathfinding via Reinforcement and Imitation Multi-Agent Learning (9)
    06.04. 2022 František Dostál: Robust Multi-Agent Path Finding and Executing (15)
    13.04. 2022 Tomáš Čelko: Trajectory Planning for Quadrotor Swarms (12)
    20.04. 2022

    Martin Bakoš: Multi-Agent Pathfinding with Real-Time Heuristic Search (8)
    27.04. 2022 Rem Lohinov: Lifelong Multi-Agent Path Finding in Large-Scale Warehouses (6)
    		  
    04.05. 2022 Jakub Mestek: Safe Multi-Agent Pathfinding with Time Uncertainty (19)
    11.05. 2022 Rektorský den  
    18.05. 2022 Matouš Mařík: On SAT-Based Approaches for Multi-Agent Path Finding with the Sum-of-Costs Objective (4) 
    Seminar ZS 2021/2022 (NAIL004):
    Středa (Wednesday) 10:40 - 12:10, lecture room S8 (Malá Strana, 1st floor)

    Seminar will run in English (Czech on demand) and it will be organized as a series of Oxford-style debates on various AI-related topics. Each side of the debate will be represented by two students (two "for" and two "against") pluse there will be a moderator. The side to which a student is allocated does not necessarily reflect his/her personal opinion!

    Debate schedule:

    • prior debate:
      • Each side is requested to send one reference (web link, paper, etc.) to the teacher by Sunday before the debate and this reference will be publicly available through the seminar web page.
    • debate day (Tuesday):
      • introduction of the topic by the moderator (5-10 minutes)
      • initial anonymous voting of audience (the result will be revealed after the discussion)
      • opening remarks (each speaker will have 5 minutes for the initial statement supporting his/her side; the order of sides is selected randomly at the beginning, speakers from both sides speak on a rota basis)
      • intra-panel discussion (between the speakers and the moderator with chance to react to the other side; 10-20 minutes)
      • Q&A (questions/comments from the audience with response from the speakers;  20-30 minutes)
      • closing remarks (each speaker will have 2 minutes; the order is reverse to the opening order)
      • final anonymous voting of audience
      • decision of the winner (the side with the increase of number of votes wins)

    Between the Oxford Debates we will include Hackathons solving selected technical problems. There is also a classroom in Microsoft Teams.

    Homework: Each student will prepare a 2-page report describing how AI can help himself/herself. Describe a problem/task that the AI system can help you to solve (for example, doing a literature review for your project). Look at existing approaches to solve that problem and suggest how the computer system could assist you in solving that problem, including ideas which techniques the system can use. The report should be like a short project propsal (this is what I want solve and why it is important; this is how it is solved so far; and this is how I suggest to solve the problem). For the report use the AAAI double-column style available at https://www.aaai.org/Publications/Templates/AuthorKit20.zip (remove the copyright from the first page).

    29.09. 2021 canceled  
    06.10. 2021 Kick-off meeting  
    13.10. 2021 Teams formation  
    20.10. 2021

    Hackathon
    Can AI help Brits (and others) with this?

    • Team A: a broad view, Uber for logistics
    • Team B: global scheduling of ships
    • Team C: local scheduling of port operations
    		  
    27.10. 2021 Oxford debate
    Should we allow AI to predict human behaviour?
    [Behaviour prediction][Crime prediction][Human Compatible AI][Anomaly detection]

    H.G., P.Š. (for); M.B., T.Č. (against)
    voting before: 86% for, 14% against
    voting after: 55% for, 45% against
    03.11. 2021 canceled (foreign trip)  
    10.11. 2021

    Oxford debate
    Will AI be ever able to understand the world by itself?
    [AI and unknown world][AI and consciousness][AI and understanding][General AI not possible?]

    		 J.H., D.N. (for); M.M., Y.S. (against)
    voting before: 44% for, 56% against
    voting after: 73% for, 27% against
    17.11. 2021 canceled (public holidays)  
    24.11. 2021 Oxford debate
    Can human fall in love with AI?
    [Asymmetrical relationship][Robot Sex][Love with hologram]

    J.K., R.P. (for); F.D., L.Z. (against)
    voting before: 90% for, 10% against
    voting after: 73% for, 27% against
    01.12. 2021

    Hackathon
    What is behind Covid waves?
    [Czech official data] [Czech social contacts] [World data] [Google mobility]

    • Team A: relations between waves (the "confirmed cases" peak is 7 days before the "hospitalisation" peak; percentage of deaths from the hospitalised is same over the time)
    • Team B: effect of government restrictions ((late) goverment restrictions decrease the peak but have no effect on the trend)
    • Team C: effects of vaccination on hospitalisations (instead of 4K it would be 12K hospitalised, if nobody is vaccinated)
    		  
    08.12. 2021 Oxford debate
    Should the user of AI be responsible for the consequences of actions done by AI?
    [Ethics of AI] [Explainable AI] [Responsible AI] [OECD Legal Instruments]    		 J.D., D.H. (for); A.K., D.R. (against)
    voting before: 73% for, 27% against
    voting after: 55% for, 45% against
    15.12. 2021 Homework  
    22.12. 2021 Homework  
    05.01. 2022 Final workshop  

     
    Seminar LS 2020/2021 (NAIL052):
    Tuesday 14:00 - 15:30, on-line via Zoom

    The seminar will start by kick-off meeting via Zoom on Tuesday, 2.3.2021, when we will discuss possible formats of the seminar and we will also decide language of seminar. We will be using a classroom in Microsoft Teams to share materials and to chat, you can register here.        

    We are solving the Flatland Challenge during the seminar. Each team will apply some approach to solve the Flatland Challenge, prepare two presentations (progress and final) for the seminar, and a final report describing the used approach and achieved results. The length of report should be 4-8 pages using the AAAI style available at https://www.aaai.org/Publications/Templates/AuthorKit20.zip (remove the copyright from the first page).

    02. 03. 2021 Kick-off meeting  
    09. 03. 2021

    Introduction to Flatland Challenge (Richard Hájek)
    followed by open discussion and teams composition
    16. 03. 2021 Teams formation  
    13. 04. 2021 Progress report, team Learn
    20. 04. 2021 Progress report, team Search
    25. 05. 2021 Final reports from each team (tentative)  
      Final report, team Learn
      Final report, team Search

     
    Seminar ZS 2020/2021 (NAIL004):
    Středa (Wednesday) 14:00 - 15:30, on-line

    Seminář začne úvodní schůzkou přes Zoom ve středu 30.9.2020, kdy se pobavíme o možných formátech semináře a také domluvíme jazyk semináře. K dispozic je také třída v prostředí Microsoft Teams.

    The seminar will start by kick-off meeting via Zoom at Wednesday, 30.9.2020, when we will discuss possible formats of the seminar and we will also decide language of seminar. There is also a classroom in Microsoft Teams.

    Seminář beží v českém jazyce přes Zoom s off-line informacemi sdílenými v prosředí Microsoft Teams. Seminář se skládá se z kombinace Oxfordských debat a hackhatonů.

    14.10. 2020

    Výzkum AI na MFF

    		  
    21.10. 2020

    Oxfordská debata

    Měli bychom na národní (evropské) úrovni regulovat výzkum umělé inteligence?

    J.M. (ano), R.S. (ne)

    Hlasování před debatou: Ano: 36%, Ne: 64%
    Hlasování po debatě: Ano: 21%, Ne: 79%
    28.10. 2020

    zrušeno

     
    04.11. 2020

    Oxfordská debata

    Je za chybu AI programu při jeho použití primárně zodpovědný tvůrce programu?

    J.U. (ano), P.V. (ne)

    Hlasování před debatou: Ano: 50%, Ne: 50%
    Hlasování po debatě: Ano: 54%, Ne: 46%
    11.11. 2020

    Hackathon

    Jak může AI pomoci při studiu?

    		  
    18.11. 2020

    Oxfordská debata

    Lze použít AI pro morální rozhodování?

    J.K. (ano), A.L. (ne)

    Hlasování před debatou: Ano: 45%, Ne: 55%
    Hlasování po debatě: Ano: 36%, Ne: 64%
    25.11. 2020

    Oxfordská debata

    Měli bychom do AI systému promítat vlastní etiku?

    M.V. (ano), T.P. (ne)

    Hlasování před debatou: Ano: 54%, Ne: 46%
    Hlasování po debatě: Ano: 79%, Ne: 21%
    02.12. 2020

    Oxfordská debata

    Měli bychom omezit prostor doporučovacímu systému na internetu? 

    J.Ř. (ano), V.V. (ne)

    Hlasování před debatou: Ano: 77%, Ne: 23%
    Hlasování po debatě: Ano: 85%, Ne: 15%
    09.12. 2020

    Oxfordská debata

    Můžeme věřit AI systémům?

    R.L. (ano), N.P. (ne)

    Hlasování před debatou: Ano: 50%, Ne: 50%
    Hlasování po debatě: Ano: 42%, Ne: 58%
    16.12. 2020

    Oxfordská debata

    Měli bychom se pokusit vytvořit obecnou umělou inteligenci?

    V.K. (ano), L.M. (ne)

    Hlasování před debatou: Ano: 85%, Ne: 15%
    Hlasování po debatě: Ano: 38%, Ne: 62%
    06.01. 2021

    Hackathon

    AI technologie pro Smart Cities


    		  

     
    Seminar LS 2019/2020 (NAIL052):
    Tuesday 09:00 - 10:30, lecture room S6 (Malá Strana, 2nd floor)

    As of March 11, 2020, all direct teaching activities are suspended. Hence instead of life presentations, each student will prepare a written report using the AAAI style available at https://www.aaai.org/Publications/Templates/AuthorKit20.zip. The length of report should be 4-8 pages.

    We will be discussing problems from multi-agent path finding (MAPF), specifically various extensions going towards practical applications. Some initial resources are available on-line:

    To get a credit from the seminar, each student will prepare a report on a specific topic based on literature (see above).

    18. 02. 2020 Opening and introduction to topics  
    25. 02. 2020 Introduction to MAPF, distribution of papers
    03. 03. 2020 MAPF on real robots
      Jan Bodnár, Scalability of MAPF  
      Matyáš Lamprecht, traffic junctions
      Adam Harmanec, warehouse
      Petra Vysušilová, learning in MAPF
      Kamil Závorka, large agents
      Karel Křesťan, MAPF in PDDL  
      Richard Pecha, Automated Parking System  
    Seminar ZS 2019/2020 (NAIL004):
    Tuesday 9:00 - 10:30, lecture room S10 (Malá Strana, 1st floor)

    Seminar will run in English (Czech on demand) and it will be organized as a series of Oxford-style debates on various AI-related topics. Each side of the debate will be represented by two students (two "for" and two "against") pluse there will be a moderator. The side to which a student is allocated does not necessarily reflect his/her personal opinion!

    Debate schedule:

    • prior debate:
      • Each side is requested to send one reference (web link, paper, etc.) to the teacher by Sunday before the debate and this reference will be publicly available through the seminar web page.
    • debate day (Tuesday):
      • introduction of the topic by the moderator (5-10 minutes)
      • initial anonymous voting of audience (the result will be revealed after the discussion)
      • opening remarks (each speaker will have 5-10 minutes for the initial statement supporting his/her side; the order of sides is selected randomly at the beginning, speakers from both sides speak on a rota basis)
      • intra-panel discussion (between the speakers and the moderator with chance to react to the other side; 10-20 minutes)
      • Q&A (questions/comments from the audience with response from the speakers;  20-30 minutes)
      • closing remarks (each speaker will have 5 minutes; the order is reverse to the opening order)
      • final anonymous voting of audience
      • decision of the winner (the side with the increase of number of votes wins)

    Possible topics of debates (send other ideas to the teacher):

    • should technical subjects care about ethical and legal issues (or should these aspects be covered by experts in social sciences)?
    • should we put all our effort to research machine learning (or should we also study other areas of AI)?
    • should AI research be done mainly in companies (or should significant portion be done by academia)?
    • should AI research be publicly regulated (or should there be no regulation or regulation of applications only)?
    • should government (or other parties) be allowed to collect and process any data (or should there be strict regulation)?
    • ...

    To get the credits from the seminar, it is required to participate in at least one debate and to attend at least 8 seminars.

    01.10. 2019

    Opening, decision about style of seminar

    		  
    08.10. 2019 Distribution and scheduling of debates
    		  
    15.10. 2019

    Should government (or other parties) be allowed to collect and process any data?

    L. Borruel, C. Caporilli (for)
    P. Březinová, Z. Šimečková (against)

    voting before: 67% for, 33% against
    voting after: 42% for, 58% against
    22.10. 2019

    Should technical subjects (AI researchers, developers) care about ethical and legal consequences of their work?

    K. Pantůčková, K. Závorka (for)
    T. De Groot, M. Strupek (against)

    voting before: 91% for, 09% against
    voting after: 38% for, 62% against
    29.10. 2019

    Should academic AI research be regulated?

    J. Musil, M. Sochor (for)
    Y. Dandarova, Y. Mu (against)

    voting before: 25% for, 75% against
    voting after: 36% for, 64% against

     
    05.11. 2019 canceled (due to foreign trip)  
    12.11. 2019 canceled (due to dean's sports day)

    19.11. 2019

    AI Research at Charles University

    • Multi-Agent Pathfinding on Real Robots (Roman Barták)
    		  
    26.11. 2019

    Should AI research be done mainly in companies?

    		 M. Belák, J. Kopál (for)
    J. Bodnár, M. Šerý (against)

    voting before: 15% for, 85% against
    voting after: 36% for, 64% against
    03.12. 2019

    Should we put most effort in AI to do research of machine learning?

    E. Düzel, K. Riahidehkord (for)
    K. Křesťan, R. Pecha (against)

    voting before: 30% for, 70% against
    voting after: 54% for, 46% against
    10.12. 2019

    AI Research at Charles University

    • Automated Planning (Roman Barták)
    • Topics of possible thesis

    		  
    17.12. 2019

    Is the risk of future AGI/Superintelligence takeover real ?

    J. Musil, M. Strupek (for)
    J. Balhar, Y. Dandarova (against)

    voting before: 11% for, 89% against
    voting after: 25% for, 75% against
    07.01. 2020 Legal and Ethical Aspect of AI (Jan Hořeňovský, Faculty of law) [PDF]
    Seminar LS 2018/2019 (NAIL052):
    Monday 12:20 - 13:50, lecture room S7 (Malá Strana, 2nd floor)

    We will be solving problems from areas such as multi-agent path finding, adversarial search, computer vision, machine learning, etc. using a swarm of Ozobot robots. The focus will be on AI (not on robotic hardware), the particular problems will be selected based on interest of each student. MAPF Scenario program by Ivan Krasičenko availabe for download. Draft of paper Multi-agent Path Finding on Real Robots in PDF. Development version of Ozoblockly that supports IR communication between robots is availabe here.

    Teams:

    ID Project Students Materials
    T1 OzoTrack (visual map reconstruction and robot tracing) Antonín Jareš vision [PDF]
    T2 OzoCar (adaptive tempomat and overpassing) Martin Strupek
    Richard Eliáš

    vision [PDF]
    T3 OzoCity (reactive path finding and collision avoidance) Jan Frauknecht
    Martin Šerý     		vision [PDF]

    Preliminary program:

    12. 02. 2019 Opening and introduction to topics  
    25. 02. 2019 Working groups forming, task distribution  
    04. 03. 2019 Initial presentatiosn of teams' visions  
    11. 03. 2019 teamproject reports  
    18. 03. 2019 homework

    25. 03. 2019

    		 homework  
    01. 04. 2019 homework  
    08. 04. 2019 progress reports  
    15. 04. 2019 homework  
    22. 04. 2019 Easter holidays  
    29. 04. 2019 homework  
    06. 05. 2019 final reports  
    13. 05. 2019 cancelled due to foreign trip  
    20. 05. 2019 cancelled due to foreign trip  
    Seminar ZS 2018/2019 (NAIL004):
    Monday 17:20 - 18:50, lecture room S8 (Malá Strana, 1st floor)

    Seminar will probably run in English (Czech on demand) and it will consist of presentations from students. Each student will present one lecture, about 60 minutes in duration plus discussion. To get the credits from the seminar, it is required to give the presentation and to attend at least 8 seminars.

    01.10. 2018

    Opening, decision about style of seminar

    		  
    08.10. 2018 Distribution and scheduling of talks
    From Abstract Models to Executable Models for Multi-agent Path Finding (Roman Barták)
    15.10. 2018 Attribute Grammars for Modeling Planning Domains(Roman Barták)
    22.10. 2018

    Multi-agent (Deep) Reinforcement Learning (Martin Pilát)
    29.10. 2018

    Swarm Robotics (Jan Kubový)
    05.11. 2018 Intelligent Intersection (Věra Škopková) video,
    12.11. 2018 canceled

    19.11. 2018

    Consciousness (Martin Červeň)
    26.11. 2018 canceled
    03.12. 2018 Autonomous Cars: Future of Driving (Igor Slovak)
    10.12. 2018 Ethics in AI (Věra Kumová)
    17.12. 2018 Generative Adversarial Networks (Aydin Ahmadli)
    Deep Stack (Tomáš Pavlín)



    07.01. 2019 canceled  
    Seminar LS 2017/2018 (NAIL052):
    Tuesday 09:00 - 10:30, lecture room S11 (Malá Strana, 1st floor)

    We will be solving problems from areas such as multi-agent path finding, adversarial search, computer vision, machine learning, etc. using a swarm of Ozobot robots. The focus will be on AI (not on robotic hardware), the particular problems will be selected based on interest of each student.

    Teams:

    ID Project Students Materials
    T1 Warehouse simulation
    (pickup and delivery)    		 David Nohejl
    Věra Škopková
    Chaman Shafiq

    vision [PDF]
    progress [PDF], video1, video2, MapsEditor,
    GitHub [TrackDesigner, OzocodeGenerator, CBS]
    final [report, presentation, video]

    T2 Manufacturing Surya Prakash Chembrolu
    Felipe Vianna
    Yuu Sakaguchi     		vision [PDF]
    progress [PDF], video, GitHub
    final [report, presentation]
    T3 Reader Alexandra Maior
    Marek Černý

    vision [PDF]
    progress [PDF]
    final [report, presentation]

    T4 Machine Learning Zdeněk Tesař vision [PDF]
    progress [PDF], demo.code
    T5 Colourly Tomáš Karella
    Tomáš Novotný
    Vladimír Patík     		vision [PDF]
    progress [PDF]

    Preliminary program:

    20. 02. 2018 Opening and introduction to topics
    27. 02. 2018 Working groups forming, task distribution  
    06. 03. 2018 project visions (teams T1,T2,T4)  
    13. 03. 2018 project visions (teams T3,T4)  
    20. 03. 2018 discussions
    27. 03. 2018 free discussions (cancelled due to a foreign trip)  
    03. 04. 2018 progress reports (team T1)  
    10. 04. 2018 progress reports (teams T2,T4)  
    17. 04. 2018 progress reports (teams T3,T5)  
    24. 04. 2018 discussions  
    01. 05. 2018 cancelled (Labour Day)  
    08. 05. 2018 cancelled (Victory Day)  
    15. 05. 2018 final reports (all teams)  
    22. 05. 2018 cancelled due to a foreign trip  

     
    Seminář ZS 2017/2018 (NAIL004):
    Tuesday 09:00 - 10:30, lecture room S1 (Malá Strana, 4th floor)

    Seminar will run in English and it will consist of presentations from students. Each student will present one lecture, about 30 minutes in duration plus discussion. To get the credits from the seminar, it is required to give the presentation and to attend at least 80% of seminars.

    03.10. 2017

    Opening, decision about style of seminar

    		  
    10.10. 2017 canceled  
    17.10. 2017 Distribution and scheduling of talks  
    24.10. 2017

    Tomáš Karella: NEAT (Neuroevolution of augmenting topologies)
    31.10. 2017

    Lukáš Meduna: Hyper NEAT
    07.11. 2017 canceled  
    14.11. 2017 Louis Outin: Text Classification by Year
    Antonín Jareš: Evolutionary Robotics

    www
    21.11. 2017

    Chaman Shafiq: Threats of AI
    28.11. 2017 Felipe Vianna: AI in Manufacturing
    Jakub Střelský: Meta-learning

    05.12. 2017 canceled  
    12.12. 2017 Marek Behún: Threats of AI
    Yuu Sakaguchi: AlphaGo
    19.12. 2017 Surya Chembrolu: DeepStack
    Karel Ha: Alpha Zero
    09.01. 2018

    Open discussion on Challenges for AI (benefits and risks of AI, philosophical foundations, ethical/moral issues)
    Some sources:
    - Rodney Brooks’ web [URL]
    - Nick Bostrom’s web [URL]
    - Moral Machine project [URL]
    - Concrete Problems in AI Safety [URL]
    - Workshop on Reliable Artificial Intelligence [URL]

    		  
    Seminar LS 2016/2017 (NAIL052):
    Tuesday 09:00 - 10:30, lecture room S6 (Malá Strana, 1. floor)

    The seminar will be devoted to application of machine learning techniques in robotics. In particular, we will focus on identifying activities of a flying drone.

    Project modules:
    Module Students Materials
    enriching sensor data Jiří Hörner
    Peter Zborovský    		 presentation [PDF], final report [PDF]
    data preprocessing Jan Pacovský presentation [PDF], final report [PDF]
    supervised learning Monika Švaralová presentation [PDF], final report [PDF]
    anomaly detection Šimon Rozsíval presentation [PDF], final report [PDF]
    learning controllers Martin Adam, Yigit Mertol Kayabasi presentation [PDF]

    [example data 1] [example data 2] [example data 3]
    [example data 4_long_flight] [example data 5_video]
    [data bebop_video] [data anomaly]
    [camera_flight_1] [camera_flight_2]
    [short_flights] 

    [complete_data_1] [complete_data_2] [complete_data_3]

    Sources:
    • Summary presentation [PDF]
    • yaDrone Recorder [URL]
    • Roman Barták, Marta Vomlelová: Using Machine Learning to Identify Activities of a Flying Drone from Sensor Readings, FLAIRS 2017 [PDF]
    • UCLA Center for Vision, Cognition, Learning, and Autonomy [URL]
    • Robot Babies (Paul Cohen at ARA) [URL]
    • Autonomous Helicopter Flight [URL]
    Seminář ZS 2016/2017 (NAIL004):
    úterý 09:00 - 10:30, posluchárna S1 (Malá Strana, 4. patro)

    Obsah a forma semináře budou domluveny na semináři 11.10.2016. V úvahu přicházejí referáty na zvolené téma dle aktuální literatury, tutoriály dle knih, řešení vlastních projektů.

    01.11. 2016

    Filip Matzner: Deep Learning, Echo State Networks, and the Edge of Chaos
    08.11. 2016 děkanský den (dean's sports day)  
    15.11. 2016 Šimon Rozsíval: Autonomous Cars
    22.11. 2016 Jan Škoda: Emotions for Artificial Agents
    29.11. 2016

    Roman Barták: Workflow models and Attribute Grammars
    06.12. 2016 odpadá (canceled)  
    13.12. 2016 Tomáš Musil: AlphaGo
    20.12. 2016 Yigit Mertol Kayabasi: Text Auto-correcting
    03.01. 2017 Filip Šedivý: Quantum AI
    10.01. 2017 Martin Černý: AI in Games  

     
    Seminář LS 2015/2016 (NAIL052):
    úterý 09:00 - 10:30, posluchárna S8 (Malá Strana, 1. patro)

    Obsah, forma a čas semináře budou domluveny na semináři 1.3.2016. V úvahu přicházejí referáty na zvolené téma dle aktuální literatury, tutoriály dle knih, řešení vlastních projektů.

     
    Seminář ZS 2015/2016 (NAIL004):
    úterý 09:00 - 10:30, posluchárna S11 (Malá Strana, 1. patro)

    Obsah a forma semináře budou domluveny na první semináři (6.10.2015). V úvahu přicházejí referáty na zvolené téma dle aktuální literatury, tutoriály dle knih, řešení vlastních projektů. Zápočet bude udělan za přednesení referátu a aktivní účast na dalších seminářích (minimálně 10 účastí).

    06.10. 2015

    Úvodní představení semináře.

    		  
    13.10. 2015 Pavel Surynek. Multi-robotické plánování.  
    20.10. 2015 Tomáš Krejčí. Distribuční strategie.  
    27.10. 2015 Natália Tyrpáková. Pravidlově-orientované rozvrhování výroby.  
    03.11. 2015 Zdeněk Tesař. Klasifikace dat z posturografických měření.
    10.11. 2015 Stuart Russell. Effective Decision Making.  
    17.11. 2015 svátek  
    24.11. 2015 Martin Mečiar. AI Mozart.  
    01.12. 2015 Tibor Baláž. Decisions in real time strategy.  
    08.12. 2015 Otakar Trunda. TBA.  
    15.12. 2015 Pavel Taufer. TBA.  
    22.12. 2015    
    05.01. 2016    
    12.01. 2016    

     
    Seminář LS 2014/2015 (NAIL052):
    pondělí 17:20 - 18:50, posluchárna S4 (Malá Strana, 3. patro)

    Seminář bude realizován formou studentských tutoriálů dle knih Synthesis Lectures on Artificial Intelligence and Machine Learning vydavatelství Morgan & Claypool Publishers (cca hodinový referát následovaný diskusí a hodnocením z pléna). Zápočet bude udělen za přednesení referátu a aktivní účast na dalších seminářích (minimálně 9 účastí).

    16. 02. 2015 Úvodní představení semináře.
    23. 02. 2015 Roman Barták: Automated Planning - A Logical Perspective
    02. 03. 2015 Roman Barták: Constraint-Based Temporal Reasoning
    09. 03. 2015 Adam Blažek: Game Theory
    16. 03. 2015 Filip Matzner: Autonomous Robotics
    23. 03. 2015 Martin Mečiar: Trading Agents  
    30. 03. 2015 Jakub Hajič: Visual Object Recognition  
    06. 04. 2015 velikonoce  
    13. 04. 2015 Promítaná přednáška (Jiří Matas: Detekce a lokalizace objektů v obrázcích) WWW
    20. 04. 2015 Neng-Fa Zhou: The Picat Language and System (guest)  
    27. 04. 2015 Tomáš Krejčí: Data Integration  
    04. 05. 2015 Michal Bilanský: Metric Learning  
    11. 05. 2015 Libor Přeučil: Inteligentní mobilní robotika: Úlohy, vize a relalita (guest)  
    18. 05. 2015 odpadá z důvodu zahraniční cesty  

     
    Seminář ZS 2014/2015 (NAIL004):
    úterý 10:40 - 12:10, posluchárna S1 (Malá Strana, 4. patro)

    Seminář bude realizován formou studentských referátů na zvolené téma (cca hodinový referát následovaný diskusí a hodnocením z pléna). Zápočet bude udělan za přednesení referátu a aktivní účast na dalších seminářích (minimálně 9 účastí). Návrhy témat k referátům:

    • Human Computation/Crowdsourcing (jak funguje řešení problémů pomocí masy lidí, www)
    • Watson (vnitřnosti Q&A systému dle IBM, www)
    • MEXAR a RAXEM (komunikace Mars Express - Země, www)
    • Remote Agent na DeepSpace 1 (jak AI řídí kosmické sondy, www, report, paper)
    • UI v JPL (jak se řídí Mars Rovers a jak satelit pozná, co má vyfotit, www)
    • Path Planning (plánování cest není jenom o cestách, www)
    • Cooperative Pathfinding (plánování cest pro více agentů najednou, www)
    • Alan Turing: Computing Machinery and Intelligence (co skrývá článek, který znamenal zrod UI, www)
    • Voting Theory (jak navrhnout správné volby, www)
    • UI v hudbě (jak rozpoznávat hudební žánry a potom je napodobovat, www)
    • Qualitative Reasoning with Time and Space (dříve nebo během, nahoře nebo uvnitř? aneb když nepotřebujeme čísla, pdf)
    • ...
    07.10. 2014

    Úvodní představení semináře.

    		  
    14.10. 2014 R. Barták, A. Vyškovský: Object Tracking by Flying Drone
    P. Surynek: Lessons Learned from the Effort to Solve Cooperative Path-Finding Optimally    		  
    21.10. 2014 F. Matzner: Short-term Motion Tracking Using Inexpensive Sensors
    28.10. 2014 svátek  
    04.11. 2014 Peter Wurman: How to Coordinate a Thousand Robots (promítaná přednáška)
    11.11. 2014

    J. Hajič: Deep Neural Networks
    18.11. 2014 odpadá z důvodu zahraniční cesty  
    25.11. 2014 odpadá z důvodu zahraniční cesty  
    02.12. 2014 J. Nádvorník: AIXI  
    09.12. 2014 Diskuse o obsahu a směrování semináře  
    16.12. 2014 T. Kadlec: Voting Theory  
    06.01. 2015 Projekt TERMES (promítaná přednáška) WWW
    video

     
    Seminář LS 2013/2014 (NAIL052):
    pondělí 14:00 - 15:30, posluchárna S6 (Malá Strana, 2. patro)

    platformNa prvním semináři budete dohodnuta struktura a obsah a semináře. Seminář bude pravděpodobně realizován formou řešení různých "výzev" a soutěží.

    Pro challenge problémy přicházejí v úvahu například následující soutěže a výzvy:

    • The Fifth Answer Set Programming Competition (www) - různé převážně "akademické" hříčky
    • MISTA 2013 challenge (www) - komplexní rozvrhovací problém
    • SAT competition (www) - různé typy problémů kódované jako problém splnitelnosti logické formule
    • International Planning Competititon (www) - velké množství plánovacích problémů
    • ICKEPS 2012 challenge (www) - téměř reálné problémy s možností nových přístupů
    • Angry Birds AI competition (www) - hra, kde si UI užije (rozpoznávání obrazu, plánování, simulace, ...)
    • a další
    • 17. 02. 2014 seznámení s náplní semináře
      24. 02. 2014 rozdělení výzev / referátů  
      03. 03. 2014 představení výzev: Car Insurance
      SAT


      10. 03. 2014 představení výzev: Automated Trading
      Co-operative Path Finding


      17. 03. 2014 představení výzev: Planning Competition
      24. 03. 2014 domácí příprava  
      31. 03. 2014 Řešící přístupy: Bio-inspired SAT
      07. 04. 2014 Řešící přístupy: Car Insurance,
      Automated Trading


      14. 04. 2014 Řešící přístupy: Planning Competition


      21 .04. 2014 velikonoce  
      28. 04. 2014 Řešící přístupy: Co-operative Path Finding
      05. 05. 2014 Promítané video: Anthony Cohn - Learning Qualitative Spatio-Temporal Activity Models  
      12. 05. 2014 Výsledky: Bio-inspired SAT


      19. 05. 2014 Výsledky:
      Automated Trading
      Car Insurance
      Co-operative Path Finding, Planning Competition
    Seminář ZS 2013/2014 (NAIL004):
    pondělí 9:00 - 10:30, posluchárna S1 (Malá Strana, 4. patro)

    Rozvrh referátů

    01.10. 2013

    Úvodní představení semináře.

    		  
    08.10. 2013

    Rozdělení referátů, promítaná přednáška (Wolfram Burgard - Probabilistic Techniques for Mobile Robot Navigation)

    		  
    15.10. 2013

    Promítaná přednáška (Vijay Kumar - Aerial Robot Swarms)

    		  
    22.10. 2013

    Jan Klátil: Volební systémy
    29.10. 2013 Jan Škoda: Úvod do Kalmanova filtru
    05.11. 2013 Milan Rybář: Cooperative Pathfinding
    12.11. 2013

    Jan Kohout: Detekce anomálií v síťovém provozu
    19.11. 2013 Milan Ježek: Qualitative Reasoning with Time and Space
    26.11. 2013 Jan Bílek: Human Computing  
    03.12. 2013

    Jakub Hajič: UI v JPL
    10.12. 2013

    Jan Tomášek, Štěpán Havránek: Robot Sumo
    17.12. 2013 Tomáš Novella: Computing Machinery and Intelligence by Alan Turing
    07.01. 2014 Marek Behún: Turingův test

     
    Seminář LS 2012/2013 (NAIL052):
    středa 17:20 - 18:50, posluchárna S4 (Malá Strana, 3. patro)

    Rozvrh referátů.

    • 20. 02. 2013 seznámení s náplní semináře
      27. 02. 2013 rozdělení výzev  
      06. 03. 2013 domácí příprava  
      13. 03. 2013 prezentace výzev (FastIron - Blue Book for Bulldozers)
      20. 03. 2013 prezentace výzev (The Marinexplore and Cornell University Whale Detection Challenge)
      27. 03. 2013 domácí příprava  
      03. 04. 2013 řešení výzev (Bulldozers)
      10. 04. 2013 řešení výzev (Whales)
      17. 04. 2013 Minulé (Petrobras)
      a budoucí výzvy (SAT)


      24 .04. 2013 Pokroky Bulldozers (a trochu Angry Birds)
      01. 05. 2013 svátek  
      08. 05. 2013 svátek  
      15. 05. 2013 Pokroky Whales


      22. 05. 2013 demo  
    Seminář ZS 2012/2013 (NAIL004):
    úterý 15:40 - 17:10, posluchárna S3 (Malá Strana, 3. patro)

    Rozvrh referátů.

    02.10. 2012

    Úvodní představení semináře, promítaná přednáška (Luis von Ahn: Duolingo: Translating the Web with Millions of People)

    		  
    09.10. 2012

    Výzkum ve skupině CSORG (Pavel Surynek: Synchronizované pohyby po grafu, Tomáš Balyo: SAT a jeho využití v plánování, Daniel Toropila: Evoluční robotika)
    pdf
    16.10. 2012

    Václav Obrázek: Kooperativní plánování cest s formacemi;
    Martin Skalský: Monte Carlo Lokalizace
    pdf
    23.10. 2012

    Marika Ivanová: Monte Carlo Tree Search,
    promítaná přednáška (Marek Michalowski: Beatbots)
    pdf
    30.10. 2012 Roman Barták: Nové techniky Induktivního logického programování
    pdf
    06.11. 2012 Rudolf Kadlec: Probabilistic model of Episodic Memory
    pdf
    13.11. 2012

    Le Khanh Chuong: Human Computation,
    Lukáš Bednařík: UI v hudbě
    20.11. 2012 Jiří Dutkevič: Any angle path planning,
    Jan Jelínek: Rozvrhování solárních panelů na ISS    		 pdf
    pdf
    27.11. 2012 Peter Hrinčár: Voting Theory,
    Filip Stočes: Remote Agent Experiment na DeepSpace 1    		 pdf
    pdf
    04.12. 2012

    Tomáš Křen: Genetické programování nad typovaným lambda kalkulem

    		pdf
    11.12. 2012

    Lenka Kovářová: Fuzzy množiny
    Tomáš Plch: Droni

    		 pdf
    pdf
    18.12. 2012 Matej Straka: Text Detection in Natural Scenes
    Štěpán Havránek, Jan Tomášek, Pavel Taufer: Hraní her
    		 pdf
    pdf
    08.01. 2013 beseda Umělá inteligence na MFF UK  

     

     
    Seminář LS 2011/2012 (NAIL052):
    pondělí 19:00 - 20:30, posluchárna S7 (Malá Strana, 2. patro)

    platformV rámci semináře byly řešeny problémy soutěže ICKEPS (International Competiton on Knowledge Engineering for Planning and Scheduling), konkrétně z následujících oblastí:

    • SACE Domain: Planning Solar Array Operations on the International Space Station
    • Petrobras Domain: Planning Ship Operations on Petroleum Platforms and Ports
    • MEX Domain: Planning Operations on the Mars Express Mission

    Podrobný popis problémů i testovací data jsou k dispozici na stránkách soutěže.

    Řešitelský tým ve složení Jan Jelínek, Martin Kolombo, Martin Pecka, Martin Hanes, Otakar Trunda, Daniel Toropila, Filip Dvořák, Roman Barták získal na soutěži ICKEPS ocenění "Outstanding Peformance on the Challening Track" a o řešení domény Petrobras vznik článek "Three Approaches to Solve the Petrobras Challenge: Exploiting Planning Techniques for Solving Real-Life Logistics Problems" na konferenci ICTAI 2012.

     
    Seminář ZS 2011/2012 (NAIL004):
    úterý 9:00 - 10:30, posluchárna S10 (Malá Strana, 1. patro)

    Rozvrh referátů.

    04.10. 2011

    Úvodní představení semináře, promítaná přednáška (David Ferrucci: Building Watson An Overview of DeepQA for the Jeopardy Challenge)

    		  
    11.10. 2011

    odpadá

    		  
    18.10. 2011

    Automated Vehicle/Urban Challenge (Jiří Harasim)
    25.10. 2011

    Automated Planning (Filip Dvořák, Daniel Toropila)
    1.11. 2011 Computer Chess (Otakar Trunda)
    08.11. 2011 přednášky hostů (posluchárna S9!)
    15.11. 2011 promítaná přednáška (Michael Wooldridge: Playing Games with Games)  
    22.11. 2011 Optimizing Limousine Service (David Marek)
    29.11. 2011 promítaná přednáška (Sumit Basu: Learning to Make Music: Interactive AI for Music Creation)  
    06.12. 2011 Bidirectional Search (Tomáš Hřebejk)
    13.12. 2011 promítaná přednáška (Raul Rojas: Autonomous Car for City Traffic)  
    20.12. 2011 How to communicate with Mars Express orbiter? (Martin Pecka)
    03.01. 2012 odpadá  
    10.01. 2012 Výzkum ve skupině CSORG. Jak se zapojit do výzkumu UI na MFF UK?