9411 An Efficient Transformer for Demosaicing via Compressed Multi-branch Attention Mechanism 1383 AN EMPIRICAL INVESTIGATION OF DOMAIN ADAPTATION ABILITY FOR CHINESE SPELLING CHECK MODELS 7060 AN EMPIRICAL STUDY ON THE IMPACT OF POSITIONAL ENCODING IN TRANSFORMER-BASED MONAURAL SPEECH ENHANCEMENT 1955 AN EN...
Deep learning is a widely applied and effective method for a broad range of applications1. Earthquake monitoring has a growing need for more efficient and robust tools for processing of increasingly large data volumes, is conceptually straightforward, and has a large quantity of available labeled dat...
Subsequently, a parameter dependency mapping graph for the transformation blocks was constructed and analyzed, and a grouping matrix F was built to record the dependency relationships among all parameters and identify parameters for simultaneous pruning. Finally, distillation was then employed to restore ...
This demonstrates the effectiveness of our proposed HFF module for efficient feature fusion. With the simultaneous introduction of the MAI and HFF modules, the Protocol#1 (MPJPE) result is reduced by 2.3 mm. In the introduction of the CIE module, the performance is further reduced by 0.4 mm, ...
To address the above two problems, we propose a hybrid transformer-based graph neural network (HTMatch) for efficient and reliable feature matching. First, we design a spatial embedding module (SEM) that embeds the spatial information across images, which enables the graph network to enhance the...
Provably (More) Sample-Efficient Offline RL with Options Xiaoyan Hu and Ho-fung Leung. NeurIPS, 2023. Double Pessimism is Provably Efficient for Distributionally Robust Offline Reinforcement Learning: Generic Algorithm and Robust Partial Coverage Jose Blanchet, Miao Lu, Tong Zhang, and Han Zhong....
Efficient semantic segmentation of large-scale point cloud scenes is a fundamental and essential task for perception or understanding the surrounding 3d environments. However, due to the vast amount of point cloud data, it is always a challenging to train deep neural networks efficiently and also di...
for pose estimation. On the other hand, the single-stage approach identifies and connects the keypoints of human bodies through keypoint detection and association analysis. The single-stage strategy does not rely on the detection of human instances, allowing for simultaneous estimation of multiple ...
for quick and efficient screening. On the other hand, the VSH mode allows for flexible pocket residue side chains during docking and conducts a two-stage conformational search on the precomputed energy grid. The first stage up-weighs the coulombic interactions threefold and runs for five iterations...
An efficient deep reinforcement learning framework for UAVs. In: 2020 21st International Symposium on Quality Electronic Design (ISQED). IEEE; 2020. p. 323–8. Bouhamed O, Ghazzai H, Besbes H, Massoud Y. Autonomous UAV navigation: a DDPG-based deep reinforcement learning approach. In: 2020 ...