1 code implementation • 29 Nov 2019 • Changlin Li, Jiefeng Peng, Liuchun Yuan, Guangrun Wang, Xiaodan Liang, Liang Lin, Xiaojun Chang
Moreover, we find that the knowledge of a network model lies not only in the network parameters but also in the network architecture.
Ranked #1 on Neural Architecture Search on CIFAR-100
1 code implementation • 12 Apr 2020 • Changlin Li, Taojiannan Yang, Sijie Zhu, Chen Chen, Shanyue Guan
Specifically, we propose a Density-Map guided object detection Network (DMNet), which is inspired from the observation that the object density map of an image presents how objects distribute in terms of the pixel intensity of the map.
1 code implementation • CVPR 2020 • Changlin Li, Jiefeng Peng, Liuchun Yuan, Guangrun Wang, Xiaodan Liang, Liang Lin, Xiaojun Chang
Remarkably, the performance of our searched architectures has exceeded the teacher model, demonstrating the practicability of our method.
1 code implementation • ICCV 2021 • Changlin Li, Tao Tang, Guangrun Wang, Jiefeng Peng, Bing Wang, Xiaodan Liang, Xiaojun Chang
In this work, we present Block-wisely Self-supervised Neural Architecture Search (BossNAS), an unsupervised NAS method that addresses the problem of inaccurate architecture rating caused by large weight-sharing space and biased supervision in previous methods.
1 code implementation • CVPR 2021 • Changlin Li, Guangrun Wang, Bing Wang, Xiaodan Liang, Zhihui Li, Xiaojun Chang
Here, we explore a dynamic network slimming regime, named Dynamic Slimmable Network (DS-Net), which aims to achieve good hardware-efficiency via dynamically adjusting filter numbers of networks at test time with respect to different inputs, while keeping filters stored statically and contiguously in hardware to prevent the extra burden.
no code implementations • 8 Aug 2021 • Rongrong Gao, Na Fan, Changlin Li, Wentao Liu, Qifeng Chen
We present a novel approach to joint depth and normal estimation for time-of-flight (ToF) sensors.
1 code implementation • ICCV 2021 • Jiefeng Peng, Jiqi Zhang, Changlin Li, Guangrun Wang, Xiaodan Liang, Liang Lin
We attribute this ranking correlation problem to the supernet training consistency shift, including feature shift and parameter shift.
1 code implementation • 21 Sep 2021 • Changlin Li, Guangrun Wang, Bing Wang, Xiaodan Liang, Zhihui Li, Xiaojun Chang
Dynamic networks have shown their promising capability in reducing theoretical computation complexity by adapting their architectures to the input during inference.
no code implementations • 17 Oct 2021 • Zutao Jiang, Changlin Li, Xiaojun Chang, Jihua Zhu, Yi Yang
Here, we present dynamic slimmable denoising network (DDS-Net), a general method to achieve good denoising quality with less computational complexity, via dynamically adjusting the channel configurations of networks at test time with respect to different noisy images.
1 code implementation • CVPR 2022 • Pengzhen Ren, Changlin Li, Guangrun Wang, Yun Xiao, Qing Du, Xiaodan Liang, Xiaojun Chang
Recently, a surge of interest in visual transformers is to reduce the computational cost by limiting the calculation of self-attention to a local window.
1 code implementation • CVPR 2022 • Changlin Li, Bohan Zhuang, Guangrun Wang, Xiaodan Liang, Xiaojun Chang, Yi Yang
First, we develop a strong manual baseline for progressive learning of ViTs, by introducing momentum growth (MoGrow) to bridge the gap brought by model growth.
1 code implementation • CVPR 2022 • Minbin Huang, Zhijian Huang, Changlin Li, Xin Chen, Hang Xu, Zhenguo Li, Xiaodan Liang
It is able to find top 0. 16\% and 0. 29\% architectures on average on two search spaces under the budget of only 50 models.
1 code implementation • CVPR 2022 • Takashi Isobe, Xu Jia, Xin Tao, Changlin Li, Ruihuang Li, Yongjie Shi, Jing Mu, Huchuan Lu, Yu-Wing Tai
Instead of directly feeding consecutive frames into a VSR model, we propose to compute the temporal difference between frames and divide those pixels into two subsets according to the level of difference.
1 code implementation • CVPR 2022 • BinBin Yang, Xinchi Deng, Han Shi, Changlin Li, Gengwei Zhang, Hang Xu, Shen Zhao, Liang Lin, Xiaodan Liang
To make ROSETTA automatically determine which experience is available and useful, a prototypical task correlation guided Gating Diversity Controller(GDC) is introduced to adaptively adjust the diversity of gates for the new task based on class-specific prototypes.
no code implementations • 28 Sep 2022 • Jiayin Cai, Changlin Li, Xin Tao, Chun Yuan, Yu-Wing Tai
This paper proposes a novel video inpainting method.
1 code implementation • 16 Oct 2022 • Tao Tang, Changlin Li, Guangrun Wang, Kaicheng Yu, Xiaojun Chang, Xiaodan Liang
Despite the success, its development and application on self-supervised vision transformers have been hindered by several barriers, including the high search cost, the lack of supervision, and the unsuitable search space.
no code implementations • 21 Nov 2022 • Changlin Li, Guangyang Wu, Yanan sun, Xin Tao, Chi-Keung Tang, Yu-Wing Tai
The learnt deformable kernel is then utilized in convolving the input frames for predicting the interpolated frame.
1 code implementation • 31 Jan 2023 • Pengzhen Ren, Changlin Li, Hang Xu, Yi Zhu, Guangrun Wang, Jianzhuang Liu, Xiaojun Chang, Xiaodan Liang
Specifically, we first propose text-to-views consistency modeling to learn correspondence for multiple views of the same input image.
no code implementations • ICCV 2023 • Kaixin Cai, Pengzhen Ren, Yi Zhu, Hang Xu, Jianzhuang Liu, Changlin Li, Guangrun Wang, Xiaodan Liang
To address this issue, we propose MixReorg, a novel and straightforward pre-training paradigm for semantic segmentation that enhances a model's ability to reorganize patches mixed across images, exploring both local visual relevance and global semantic coherence.
1 code implementation • 22 Aug 2023 • Xueyi Liu, Rui Hou, Yanglei Gan, Da Luo, Changlin Li, Xiaojun Shi, Qiao Liu
In addition, we design a multi-perspective attention mechanism that align relevant opinion information with respect to the given aspect.
no code implementations • ICCV 2023 • Xinchi Deng, Han Shi, Runhui Huang, Changlin Li, Hang Xu, Jianhua Han, James Kwok, Shen Zhao, Wei zhang, Xiaodan Liang
Compared with the existing methods, GrowCLIP improves 2. 3% average top-1 accuracy on zero-shot image classification of 9 downstream tasks.
1 code implementation • 9 Oct 2023 • Xuwei Xu, Changlin Li, Yudong Chen, Xiaojun Chang, Jiajun Liu, Sen Wang
By allowing the idle tokens to be re-selected in the following layers, IdleViT mitigates the negative impact of improper pruning in the early stages.
1 code implementation • 2 Mar 2024 • Guangrun Wang, Changlin Li, Liuchun Yuan, Jiefeng Peng, Xiaoyu Xian, Xiaodan Liang, Xiaojun Chang, Liang Lin
Addressing this problem, we modularize a large search space into blocks with small search spaces and develop a family of models with the distilling neural architecture (DNA) techniques.
no code implementations • 10 Apr 2024 • Guangyang Wu, Xin Tao, Changlin Li, Wenyi Wang, Xiaohong Liu, Qingqing Zheng
In practice, motion estimates often prove to be error-prone, resulting in misaligned features.