Generative Hybrid Representations for Activity Forecasting with No-Regret Learning

被引:48
|
作者
Guan, Jiaqi [1 ,2 ]
Yuan, Ye [1 ]
Kitani, Kris M. [1 ]
Rhinehart, Nicholas [1 ,3 ]
机构
[1] Carnegie Mellon Univ, Pittsburgh, PA 15213 USA
[2] Univ Illinois, Champaign, IL 61820 USA
[3] Univ Calif Berkeley, Berkeley, CA USA
关键词
D O I
10.1109/CVPR42600.2020.00025
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Automatically reasoning about future human behaviors is a difficult problem but has significant practical applications to assistive systems. Part of this difficulty stems from learning systems' inability to represent all kinds of behaviors. Some behaviors, such as motion, are best described with continuous representations, whereas others, such as picking up a cup, are best described with discrete representations. Furthermore, human behavior is generally not fixed: people can change their habits and routines. This suggests these systems must be able to learn and adapt continuously. In this work, we develop an efficient deep generative model to jointly forecast a person's future discrete actions and continuous motions. On a large-scale egocentric dataset, EPIC-KITCHENS, we observe our method generates high-quality and diverse samples while exhibiting better generalization than related generative models. Finally, we propose a variant to continually learn our model from streaming data, observe its practical effectiveness, and theoretically justify its learning efficiency.
引用
收藏
页码:170 / 179
页数:10
相关论文
共 50 条
  • [21] Robustness enhancement of complex networks via No-Regret learning
    Sohn, Insoo
    ICT EXPRESS, 2019, 5 (03): : 163 - 166
  • [22] No-Regret Reinforcement Learning with Heavy-Tailed Rewards
    Zhuang, Vincent
    Sui, Yanan
    24TH INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS (AISTATS), 2021, 130
  • [23] No-Regret Learning in Collaborative Spectrum Sensing with Malicious Nodes
    Zhu, Quanyan
    Han, Zhu
    Basar, Tamer
    2010 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, 2010,
  • [24] No-Regret Learning with Unbounded Losses: The Case of Logarithmic Pooling
    Neyman, Eric
    Roughgarden, Tim
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 36 (NEURIPS 2023), 2023,
  • [25] On the Impossibility of Convergence of Mixed Strategies with Optimal No-Regret Learning
    Muthukumar, Vidya
    Phade, Soham
    Sahai, Anant
    MATHEMATICS OF OPERATIONS RESEARCH, 2024,
  • [26] No-regret learning for repeated concave games with lossy bandits
    Liu, Wenting
    Lei, Jinlong
    Yi, Peng
    2021 60TH IEEE CONFERENCE ON DECISION AND CONTROL (CDC), 2021, : 936 - 941
  • [27] Near-Optimal No-Regret Learning in General Games
    Daskalakis, Constantinos
    Fishelson, Maxwell
    Golowich, Noah
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34
  • [28] No-Regret Online Reinforcement Learning with Adversarial Losses and Transitions
    Jin, Tiancheng
    Liu, Junyan
    Rouyer, Chloe
    Chang, William
    Wei, Chen-Yu
    Luo, Haipeng
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 36 (NEURIPS 2023), 2023,
  • [29] No-Regret Learning and Mixed Nash Equilibria: They Do Not Mix
    Flokas, Lampros
    Vlatakis-Gkaragkounis, Emmanouil V.
    Lianeas, Thanasis
    Mertikopoulos, Panayotis
    Piliouras, Georgios
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020, 2020, 33
  • [30] Tighter Robust Upper Bounds for Options via No-Regret Learning
    Xue, Shan
    Du, Ye
    Xu, Liang
    THIRTY-SEVENTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 37 NO 4, 2023, : 5348 - 5356