Zobrazeno 1 - 10
of 4 375
pro vyhledávání: '"Li, HongLiang"'
Autor:
Xu, Linfeng, Meng, Fanman, Wu, Qingbo, Pan, Lili, Qiu, Heqian, Wang, Lanxiao, Chen, Kailong, Geng, Kanglei, Qian, Yilei, Wang, Haojie, Zhou, Shuchang, Ling, Shimou, Liu, Zejia, Chen, Nanlin, Xu, Yingjie, Cheng, Shaoxu, Tan, Bowen, Xu, Ziyong, Li, Hongliang
The application of activity recognition in the ``AI + Education" field is gaining increasing attention. However, current work mainly focuses on the recognition of activities in manually captured videos and a limited number of activity types, with lit
Externí odkaz:
http://arxiv.org/abs/2410.12337
In this paper, we explore a novel Text-supervised Egocentic Semantic Segmentation (TESS) task that aims to assign pixel-level categories to egocentric images weakly supervised by texts from image-level labels. In this task with prospective potential,
Externí odkaz:
http://arxiv.org/abs/2410.01341
Recent advancements in prompt tuning have successfully adapted large-scale models like Contrastive Language-Image Pre-trained (CLIP) for downstream tasks such as scene text detection. Typically, text prompt complements the text encoder's input, focus
Externí odkaz:
http://arxiv.org/abs/2409.13576
Autor:
Qian, Yilei, Geng, Kanglei, Chen, Kailong, Cheng, Shaoxu, Xu, Linfeng, Li, Hongliang, Meng, Fanman, Wu, Qingbo
The application of activity recognition in the "AI + Education" field is gaining increasing attention. However, current work mainly focuses on the recognition of activities in manually captured videos and a limited number of activity types, with litt
Externí odkaz:
http://arxiv.org/abs/2409.03354
We introduce a novel TRUNC finite element in n dimensions, encompassing the traditional TRUNC triangle as a particular instance. By establishing the weak continuity identity, we identify it as crucial for error estimate. This element is utilized to a
Externí odkaz:
http://arxiv.org/abs/2409.00748
Text-rich document understanding (TDU) refers to analyzing and comprehending documents containing substantial textual content. With the rapid evolution of large language models (LLMs), they have been widely leveraged for TDU due to their remarkable v
Externí odkaz:
http://arxiv.org/abs/2408.15045
Distribution-Level Memory Recall for Continual Learning: Preserving Knowledge and Avoiding Confusion
Autor:
Cheng, Shaoxu, Geng, Kanglei, He, Chiyuan, Qiu, Zihuan, Xu, Linfeng, Qiu, Heqian, Wang, Lanxiao, Wu, Qingbo, Meng, Fanman, Li, Hongliang
Continual Learning (CL) aims to enable Deep Neural Networks (DNNs) to learn new data without forgetting previously learned knowledge. The key to achieving this goal is to avoid confusion at the feature level, i.e., avoiding confusion within old tasks
Externí odkaz:
http://arxiv.org/abs/2408.02695
Autor:
Chen, Shuai, Meng, Fanman, Wu, Chenhao, Wei, Haoran, Zhang, Runtong, Wu, Qingbo, Xu, Linfeng, Li, Hongliang
Few-Shot Segmentation (FSS) aims to segment novel classes using only a few annotated images. Despite considerable process under pixel-wise support annotation, current FSS methods still face three issues: the inflexibility of backbone upgrade without
Externí odkaz:
http://arxiv.org/abs/2407.16182
Autor:
Huang, Kaiyu, Mo, Fengran, Li, Hongliang, Li, You, Zhang, Yuanchi, Yi, Weijian, Mao, Yulong, Liu, Jinchen, Xu, Yuzhuang, Xu, Jinan, Nie, Jian-Yun, Liu, Yang
The rapid development of Large Language Models (LLMs) demonstrates remarkable multilingual capabilities in natural language processing, attracting global attention in both academia and industry. To mitigate potential discrimination and enhance the ov
Externí odkaz:
http://arxiv.org/abs/2405.10936
Autor:
Wu, Chenhao, Wu, Qingbo, Wei, Haoran, Chen, Shuai, Wang, Lei, Ngan, King Ngi, Meng, Fanman, Li, Hongliang
Despite demonstrating superior rate-distortion (RD) performance, learning-based image compression (LIC) algorithms have been found to be vulnerable to malicious perturbations in recent studies. However, the adversarial attacks considered in existing
Externí odkaz:
http://arxiv.org/abs/2405.07717