Zobrazeno 41 - 50
of 131
pro vyhledávání: '"Birchfield, Stan"'
Autor:
Tseng, Hung-Yu, De Mello, Shalini, Tremblay, Jonathan, Liu, Sifei, Birchfield, Stan, Yang, Ming-Hsuan, Kautz, Jan
Viewpoint estimation for known categories of objects has been improved significantly thanks to deep networks and large datasets, but generalization to unknown categories is still very challenging. With an aim towards improving performance on unknown
Externí odkaz:
http://arxiv.org/abs/1905.04957
Autor:
Tang, Zheng, Naphade, Milind, Liu, Ming-Yu, Yang, Xiaodong, Birchfield, Stan, Wang, Shuo, Kumar, Ratnesh, Anastasiu, David, Hwang, Jenq-Neng
Urban traffic optimization using traffic cameras as sensors is driving the need to advance state-of-the-art multi-target multi-camera (MTMC) tracking. This work introduces CityFlow, a city-scale traffic camera dataset consisting of more than 3 hours
Externí odkaz:
http://arxiv.org/abs/1903.09254
Autor:
Cheng, Ching-An, Mukadam, Mustafa, Issac, Jan, Birchfield, Stan, Fox, Dieter, Boots, Byron, Ratliff, Nathan
We develop a novel policy synthesis algorithm, RMPflow, based on geometrically consistent transformations of Riemannian Motion Policies (RMPs). RMPs are a class of reactive motion policies designed to parameterize non-Euclidean behaviors as dynamical
Externí odkaz:
http://arxiv.org/abs/1811.07049
Autor:
Prakash, Aayush, Boochoon, Shaad, Brophy, Mark, Acuna, David, Cameracci, Eric, State, Gavriel, Shapira, Omer, Birchfield, Stan
We present structured domain randomization (SDR), a variant of domain randomization (DR) that takes into account the structure and context of the scene. In contrast to DR, which places objects and distractors randomly according to a uniform probabili
Externí odkaz:
http://arxiv.org/abs/1810.10093
Autor:
Sundaralingam, Balakumar, Lambert, Alexander, Handa, Ankur, Boots, Byron, Hermans, Tucker, Birchfield, Stan, Ratliff, Nathan, Fox, Dieter
Current methods for estimating force from tactile sensor signals are either inaccurate analytic models or task-specific learned models. In this paper, we explore learning a robust model that maps tactile sensor signals to force. We specifically explo
Externí odkaz:
http://arxiv.org/abs/1810.06187
Autor:
Tremblay, Jonathan, To, Thang, Sundaralingam, Balakumar, Xiang, Yu, Fox, Dieter, Birchfield, Stan
Using synthetic data for training deep neural networks for robotic manipulation holds the promise of an almost unlimited amount of pre-labeled training data, generated safely out of harm's way. One of the key challenges of synthetic data, to date, ha
Externí odkaz:
http://arxiv.org/abs/1809.10790
Learning a policy capable of moving an agent between any two states in the environment is important for many robotics problems involving navigation and manipulation. Due to the sparsity of rewards in such tasks, applying reinforcement learning in the
Externí odkaz:
http://arxiv.org/abs/1807.01425
Autor:
Tremblay, Jonathan, To, Thang, Molchanov, Artem, Tyree, Stephen, Kautz, Jan, Birchfield, Stan
We present a system to infer and execute a human-readable program from a real-world demonstration. The system consists of a series of neural networks to perform perception, program generation, and program execution. Leveraging convolutional pose mach
Externí odkaz:
http://arxiv.org/abs/1805.07054
We present a new dataset, called Falling Things (FAT), for advancing the state-of-the-art in object detection and 3D pose estimation in the context of robotics. By synthetically combining object models and backgrounds of complex composition and high
Externí odkaz:
http://arxiv.org/abs/1804.06534
Autor:
Tremblay, Jonathan, Prakash, Aayush, Acuna, David, Brophy, Mark, Jampani, Varun, Anil, Cem, To, Thang, Cameracci, Eric, Boochoon, Shaad, Birchfield, Stan
We present a system for training deep neural networks for object detection using synthetic images. To handle the variability in real-world data, the system relies upon the technique of domain randomization, in which the parameters of the simulator$-$
Externí odkaz:
http://arxiv.org/abs/1804.06516