Zobrazeno 1 - 10
of 211
pro vyhledávání: '"Lan Xuguang"'
This paper presents a general scheme for enhancing the convergence and performance of DETR (DEtection TRansformer). We investigate the slow convergence problem in transformers from a new perspective, suggesting that it arises from the self-attention
Externí odkaz:
http://arxiv.org/abs/2407.11699
Effective exploration is crucial to discovering optimal strategies for multi-agent reinforcement learning (MARL) in complex coordination tasks. Existing methods mainly utilize intrinsic rewards to enable committed exploration or use role-based learni
Externí odkaz:
http://arxiv.org/abs/2402.17978
Linguistic ambiguity is ubiquitous in our daily lives. Previous works adopted interaction between robots and humans for language disambiguation. Nevertheless, when interactive robots are deployed in daily environments, there are significant challenge
Externí odkaz:
http://arxiv.org/abs/2402.11792
Interactive visual grounding in Human-Robot Interaction (HRI) is challenging yet practical due to the inevitable ambiguity in natural languages. It requires robots to disambiguate the user input by active information gathering. Previous approaches of
Externí odkaz:
http://arxiv.org/abs/2401.16699
Autor:
Jiang, Zhenhao, Zeng, Biao, Feng, Hao, Liu, Jin, Fan, Jicong, Zhang, Jie, Jia, Jia, Hu, Ning, Chen, Xingyu, Lan, Xuguang
Large-scale online recommender system spreads all over the Internet being in charge of two basic tasks: Click-Through Rate (CTR) and Post-Click Conversion Rate (CVR) estimations. However, traditional CVR estimators suffer from well-known Sample Selec
Externí odkaz:
http://arxiv.org/abs/2307.09193
Manipulation relationship detection (MRD) aims to guide the robot to grasp objects in the right order, which is important to ensure the safety and reliability of grasping in object stacked scenes. Previous works infer manipulation relationship by dee
Externí odkaz:
http://arxiv.org/abs/2304.12592
In scenarios involving the grasping of multiple targets, the learning of stacking relationships between objects is fundamental for robots to execute safely and efficiently. However, current methods lack subdivision for the hierarchy of stacking relat
Externí odkaz:
http://arxiv.org/abs/2303.07828
Due to the representation limitation of the joint Q value function, multi-agent reinforcement learning methods with linear value decomposition (LVD) or monotonic value decomposition (MVD) suffer from relative overgeneralization. As a result, they can
Externí odkaz:
http://arxiv.org/abs/2211.12075
Robotic Grasping has always been an active topic in robotics since grasping is one of the fundamental but most challenging skills of robots. It demands the coordination of robotic perception, planning, and control for robustness and intelligence. How
Externí odkaz:
http://arxiv.org/abs/2202.03631
Publikováno v:
In Neurocomputing 28 September 2024 599