2 code implementations • 27 Mar 2024 • Yanwei Li, Yuechen Zhang, Chengyao Wang, Zhisheng Zhong, Yixin Chen, Ruihang Chu, Shaoteng Liu, Jiaya Jia
We try to narrow the gap by mining the potential of VLMs for better performance and any-to-any workflow from three aspects, i. e., high-resolution visual tokens, high-quality data, and VLM-guided generation.
Ranked #9 on Visual Question Answering on MM-Vet
4 code implementations • 23 May 2023 • Jiequan Cui, Zhuotao Tian, Zhisheng Zhong, Xiaojuan Qi, Bei Yu, Hanwang Zhang
In this paper, we delve deeper into the Kullback-Leibler (KL) Divergence loss and observe that it is equivalent to the Doupled Kullback-Leibler (DKL) Divergence loss that consists of 1) a weighted Mean Square Error (wMSE) loss and 2) a Cross-Entropy loss incorporating soft labels.
2 code implementations • CVPR 2023 • Zhisheng Zhong, Jiequan Cui, Yibo Yang, Xiaoyang Wu, Xiaojuan Qi, Xiangyu Zhang, Jiaya Jia
Based on our empirical and theoretical analysis, we point out that semantic segmentation naturally brings contextual correlation and imbalanced distribution among classes, which breaks the equiangular and maximally separated structure of neural collapse for both feature centers and classifiers.
4 code implementations • 26 Sep 2022 • Jiequan Cui, Zhisheng Zhong, Zhuotao Tian, Shu Liu, Bei Yu, Jiaya Jia
Based on theoretical analysis, we observe that supervised contrastive loss tends to bias high-frequency classes and thus increases the difficulty of imbalanced learning.
Ranked #7 on Long-tail Learning on iNaturalist 2018
5 code implementations • 5 Apr 2022 • Jiequan Cui, Yuhui Yuan, Zhisheng Zhong, Zhuotao Tian, Han Hu, Stephen Lin, Jiaya Jia
In this paper, we study the problem of class imbalance in semantic segmentation.
Ranked #21 on Semantic Segmentation on ADE20K
2 code implementations • 22 Mar 2022 • Zhisheng Zhong, Jiequan Cui, Zeming Li, Eric Lo, Jian Sun, Jiaya Jia
Given the promising performance of contrastive learning, we propose Rebalanced Siamese Contrastive Mining (ResCom) to tackle imbalanced recognition.
Ranked #5 on Long-tail Learning on CIFAR-10-LT (ρ=10)
5 code implementations • ICCV 2021 • Jiequan Cui, Zhisheng Zhong, Shu Liu, Bei Yu, Jiaya Jia
In this paper, we propose Parametric Contrastive Learning (PaCo) to tackle long-tailed recognition.
Ranked #14 on Long-tail Learning on iNaturalist 2018
5 code implementations • CVPR 2021 • Zhisheng Zhong, Jiequan Cui, Shu Liu, Jiaya Jia
Motivated by the fact that predicted probability distributions of classes are highly related to the numbers of class instances, we propose label-aware smoothing to deal with different degrees of over-confidence for classes and improve classifier learning.
Ranked #16 on Long-tail Learning on CIFAR-10-LT (ρ=100)
5 code implementations • 26 Jan 2021 • Jiequan Cui, Shu Liu, Zhuotao Tian, Zhisheng Zhong, Jiaya Jia
From this perspective, the trivial solution utilizes different branches for the head, medium, and tail classes respectively, and then sums their outputs as the final results is not feasible.
Ranked #22 on Long-tail Learning on CIFAR-10-LT (ρ=10)
1 code implementation • 15 Jul 2020 • Zhisheng Zhong, Hiroaki Akutsu, Kiyoharu Aizawa
In this paper, we propose a channel-level variable quantization network to dynamically allocate more bitrates for significant channels and withdraw bitrates for negligible channels.
5 code implementations • ICCV 2019 • Xia Li, Zhisheng Zhong, Jianlong Wu, Yibo Yang, Zhouchen Lin, Hong Liu
It is designed to compute the representation of each position by a weighted sum of the features at all positions.
Ranked #11 on Semantic Segmentation on COCO-Stuff test
no code implementations • 18 Jun 2019 • Zhisheng Zhong, Fangyin Wei, Zhouchen Lin, Chao Zhang
Furthermore, we propose that weight tensors in networks with proper order and balanced dimension are easier to be compressed.
1 code implementation • 15 May 2019 • Xingyu Xie, Jianlong Wu, Zhisheng Zhong, Guangcan Liu, Zhouchen Lin
Recently, a number of learning-based optimization methods that combine data-driven architectures with the classical optimization algorithms have been proposed and explored, showing superior empirical performance in solving various ill-posed inverse problems, but there is still a scarcity of rigorous analysis about the convergence behaviors of learning-based optimization.
no code implementations • NeurIPS 2018 • Zhisheng Zhong, Tiancheng Shen, Yibo Yang, Zhouchen Lin, Chao Zhang
To solve these problems, we propose the Super-Resolution CliqueNet (SRCliqueNet) to reconstruct the high resolution (HR) image with better textural details in the wavelet domain.
3 code implementations • CVPR 2018 • Yibo Yang, Zhisheng Zhong, Tiancheng Shen, Zhouchen Lin
In contrast to prior networks, there are both forward and backward connections between any two layers in the same block.