Learning Active Camera for Multi-Object Navigation
28. listopadu 2022
Řečníci
Peihao Chen
Řečník · 0 sledujících
Kunyang Lin
Řečník · 0 sledujících
Wenbing Huang
Řečník · 0 sledujících
Mingkui Tan
Řečník · 0 sledujících
Chuang Gan
Řečník · 0 sledujících
O prezentaci
Getting robots to navigate to multiple objects autonomously is essential yet difficult in robot applications. One of the key challenges is how to explore environments efficiently with camera sensors only. Existing navigation methods mainly focus on fixed cameras and few attempts have been made to navigate with active cameras. As a result, the agent may take a very long time to perceive the environment due to limited camera scope. In contrast, humans typically gain a larger field of view by looking around for a better perception of the environment. How to make robots perceive the environment as efficiently as humans is a fundamental problem in robotics. In this paper, we consider navigating to multiple objects more efficiently with active cameras. Specifically, we cast moving camera to a Markov Decision Process and reformulate the active camera problem as a reinforcement learning problem. However, we have to address two new challenges: 1) how to learn a good camera policy in complex environments and 2) how to coordinate it with the navigation policy. To address these, we carefully design a reward function to encourage the agent to explore more areas by moving camera actively. Moreover, we exploit human experience to infer a rule-based camera action to guide the learning process. Last, to better coordinate two kinds of policies, the camera policy takes navigation actions into account when making camera moving decisions. Experimental results show our camera policy consistently improves the performance of multi-object navigation over four baselines on two datasets.Getting robots to navigate to multiple objects autonomously is essential yet difficult in robot applications. One of the key challenges is how to explore environments efficiently with camera sensors only. Existing navigation methods mainly focus on fixed cameras and few attempts have been made to navigate with active cameras. As a result, the agent may take a very long time to perceive the environment due to limited camera scope. In contrast, humans typically gain a larger field of view by looki…
Organizátor
NeurIPS 2022
Účet · 962 sledujících
Baví vás formát? Nechte SlidesLive zachytit svou akci!
Profesionální natáčení a streamování po celém světě.
Sdílení
Doporučená videa
Prezentace na podobné téma, kategorii nebo přednášejícího
Robust Scheduling with GFlowNets
Zhlédnout později
Pro uložení prezentace do věčného trezoru hlasovalo 0 diváků, což je 0.0 %
P2P: Tuning Pre-trained Image Models for Point Cloud Analysis with Point-to-Pixel Prompting
Zhlédnout později
Ziyi Wang, …
Pro uložení prezentace do věčného trezoru hlasovalo 0 diváků, což je 0.0 %
Uncertainty Disentanglement with Non-stationary Heteroscedastic Gaussian Processes for Active Learning
Zhlédnout později
Zeel B Patel, …
Pro uložení prezentace do věčného trezoru hlasovalo 0 diváků, což je 0.0 %
Optimal Binary Classification Beyond Accuracy
Zhlédnout později
Pro uložení prezentace do věčného trezoru hlasovalo 0 diváků, což je 0.0 %
HUMANISE: Language-conditioned Human Motion Generation in 3D Scenes
Zhlédnout později
Zan Wang, …
Pro uložení prezentace do věčného trezoru hlasovalo 0 diváků, což je 0.0 %
Chartalist: Labeled Graph Datasets for UTXO and Account-based Blockchains
Zhlédnout později
Pro uložení prezentace do věčného trezoru hlasovalo 0 diváků, což je 0.0 %