Zobrazeno 1 - 10
of 109
pro vyhledávání: '"Yamins, Daniel L. K."'
Autor:
Long, Bria, Xiang, Violet, Stojanov, Stefan, Sparks, Robert Z., Yin, Zi, Keene, Grace E., Tan, Alvin W. M., Feng, Steven Y., Zhuang, Chengxu, Marchman, Virginia A., Yamins, Daniel L. K., Frank, Michael C.
Human children far exceed modern machine learning algorithms in their sample efficiency, achieving high performance in key domains with much less data than current models. This ''data gap'' is a key challenge both for building intelligent artificial
Externí odkaz:
http://arxiv.org/abs/2406.10447
Autor:
Venkatesh, Rahul, Chen, Honglin, Feigelis, Kevin, Bear, Daniel M., Jedoui, Khaled, Kotar, Klemen, Binder, Felix, Lee, Wanhee, Liu, Sherry, Smith, Kevin A., Fan, Judith E., Yamins, Daniel L. K.
The ability to understand physical dynamics is critical for agents to act in the world. Here, we use Counterfactual World Modeling (CWM) to extract vision structures for dynamics understanding. CWM uses a temporally-factored masking policy for masked
Externí odkaz:
http://arxiv.org/abs/2312.06721
The human visual system can effortlessly recognize an object under different extrinsic factors such as lighting, object poses, and background, yet current computer vision systems often struggle with these variations. An important step to understandin
Externí odkaz:
http://arxiv.org/abs/2311.00750
Autor:
Bear, Daniel M., Feigelis, Kevin, Chen, Honglin, Lee, Wanhee, Venkatesh, Rahul, Kotar, Klemen, Durango, Alex, Yamins, Daniel L. K.
Leading approaches in machine vision employ different architectures for different tasks, trained on costly task-specific labeled datasets. This complexity has held back progress in areas, such as robotics, where robust task-general perception remains
Externí odkaz:
http://arxiv.org/abs/2306.01828
Autor:
Martinez, Julio, Binder, Felix, Wang, Haoliang, Haber, Nick, Fan, Judith, Yamins, Daniel L. K.
Humans are interactive agents driven to seek out situations with interesting physical dynamics. Here we formalize the functional form of physical intrinsic motivation. We first collect ratings of how interesting humans find a variety of physics scena
Externí odkaz:
http://arxiv.org/abs/2305.13452
Infants explore their complex physical and social environment in an organized way. To gain insight into what intrinsic motivations may help structure this exploration, we create a virtual infant agent and place it in a developmentally-inspired 3D env
Externí odkaz:
http://arxiv.org/abs/2305.13396
Autor:
Chen, Honglin, Venkatesh, Rahul, Friedman, Yoni, Wu, Jiajun, Tenenbaum, Joshua B., Yamins, Daniel L. K., Bear, Daniel M.
Self-supervised, category-agnostic segmentation of real-world images is a challenging open problem in computer vision. Here, we show how to learn static grouping priors from motion self-supervision by building on the cognitive science concept of a Sp
Externí odkaz:
http://arxiv.org/abs/2205.08515
Autor:
Kunin, Daniel, Sagastuy-Brena, Javier, Gillespie, Lauren, Margalit, Eshed, Tanaka, Hidenori, Ganguli, Surya, Yamins, Daniel L. K.
Publikováno v:
Neural Computation (2024) 36 (1) 151-174
In this work we explore the limiting dynamics of deep neural networks trained with stochastic gradient descent (SGD). As observed previously, long after performance has converged, networks continue to move through parameter space by a process of anom
Externí odkaz:
http://arxiv.org/abs/2107.09133
Autor:
Bear, Daniel M., Wang, Elias, Mrowca, Damian, Binder, Felix J., Tung, Hsiao-Yu Fish, Pramod, R. T., Holdaway, Cameron, Tao, Sirui, Smith, Kevin, Sun, Fan-Yun, Fei-Fei, Li, Kanwisher, Nancy, Tenenbaum, Joshua B., Yamins, Daniel L. K., Fan, Judith E.
While current vision algorithms excel at many challenging tasks, it is unclear how well they understand the physical dynamics of real-world environments. Here we introduce Physion, a dataset and benchmark for rigorously evaluating the ability to pred
Externí odkaz:
http://arxiv.org/abs/2106.08261
Autor:
Gan, Chuang, Zhou, Siyuan, Schwartz, Jeremy, Alter, Seth, Bhandwaldar, Abhishek, Gutfreund, Dan, Yamins, Daniel L. K., DiCarlo, James J, McDermott, Josh, Torralba, Antonio, Tenenbaum, Joshua B.
We introduce a visually-guided and physics-driven task-and-motion planning benchmark, which we call the ThreeDWorld Transport Challenge. In this challenge, an embodied agent equipped with two 9-DOF articulated arms is spawned randomly in a simulated
Externí odkaz:
http://arxiv.org/abs/2103.14025