no code implementations • ECCV 2020 • Fang Zhao, Shengcai Liao, Guo-Sen Xie, Jian Zhao, Kaihao Zhang, Ling Shao
On the other hand, mutual instance selection further selects reliable and informative instances for training according to the peer-confidence and relationship disagreement of the networks.
no code implementations • ECCV 2020 • Xiaobo Wang, Tianyu Fu, Shengcai Liao, Shuo Wang, Zhen Lei, Tao Mei
Knowledge distillation is an effective tool to compress large pre-trained Convolutional Neural Networks (CNNs) or their ensembles into models applicable to mobile and embedded devices.
1 code implementation • 25 Apr 2024 • Jiehui Huang, Xiao Dong, Wenhui Song, Hanhui Li, Jun Zhou, Yuhao Cheng, Shutao Liao, Long Chen, Yiqiang Yan, Shengcai Liao, Xiaodan Liang
ConsistentID comprises two key components: a multimodal facial prompt generator that combines facial features, corresponding facial descriptions and the overall facial context to enhance precision in facial details, and an ID-preservation network optimized through the facial attention localization strategy, aimed at preserving ID consistency in facial regions.
no code implementations • 15 Feb 2024 • Zehao Xiao, Jiayi Shen, Mohammad Mahdi Derakhshani, Shengcai Liao, Cees G. M. Snoek
To effectively encode the distribution information and their relationships, we further introduce a transformer inference network with a pseudo-shift training mechanism.
1 code implementation • NeurIPS 2023 • Yingjun Du, Zehao Xiao, Shengcai Liao, Cees Snoek
Furthermore, we introduce a task-guided diffusion process within the prototype space, enabling the meta-learning of a generative process that transitions from a vanilla prototype to an overfitted prototype.
1 code implementation • 31 May 2023 • Guian Fang, Zutao Jiang, Jianhua Han, Guansong Lu, Hang Xu, Shengcai Liao, Xiaodan Liang
Recent advances in text-to-image diffusion models have achieved remarkable success in generating high-quality, realistic images from textual descriptions.
no code implementations • 18 Apr 2023 • Rongliang Wu, Yingchen Yu, Fangneng Zhan, Jiahui Zhang, Shengcai Liao, Shijian Lu
POCE achieves the more accessible and realistic pose-controllable expression editing by mapping face images into UV space, where facial expressions and head poses can be disentangled and edited separately.
no code implementations • CVPR 2023 • Kaiwen Cui, Yingchen Yu, Fangneng Zhan, Shengcai Liao, Shijian Lu1, Eric Xing
The first is aggregated generative KD that mitigates the discriminator overfitting by challenging the discriminator with harder learning tasks and distilling more generalizable knowledge from the pre-trained models.
no code implementations • 10 Mar 2023 • Xinghong Liu, Yi Zhou, Tao Zhou, Jie Qin, Shengcai Liao
Open-set domain adaptation aims to not only recognize target samples belonging to common classes shared by source and target domains but also perceive unknown class samples.
1 code implementation • 22 Feb 2023 • Zehao Xiao, XianTong Zhen, Shengcai Liao, Cees G. M. Snoek
In this paper, we propose energy-based sample adaptation at test time for domain generalization.
no code implementations • 14 Jul 2022 • Xingping Dong, Shengcai Liao, Bo Du, Ling Shao
Most existing few-shot learning (FSL) methods require a large amount of labeled data in meta-training, which is a major limit.
no code implementations • 8 Jul 2022 • Jinpeng Li, Haibo Jin, Shengcai Liao, Ling Shao, Pheng-Ann Heng
This paper presents a Refinement Pyramid Transformer (RePFormer) for robust facial landmark detection.
2 code implementations • CVPR 2022 • Yanan Wang, Xuezhi Liang, Shengcai Liao
To address this, in this work, an automatic approach is proposed to directly clone the whole outfits from real-world person images to virtual 3D characters, such that any virtual person thus created will appear very similar to its real-world counterpart.
Ranked #1 on Unsupervised Domain Adaptation on ClonedPerson (using extra training data)
Generalizable Person Re-identification Unsupervised Domain Adaptation +1
2 code implementations • 10 Jan 2022 • Irtiza Hasan, Shengcai Liao, Jinpeng Li, Saad Ullah Akram, Ling Shao
As for the data, we show that the autonomous driving benchmarks are monotonous in nature, that is, they are not diverse in scenarios and dense in pedestrians.
no code implementations • 29 Nov 2021 • Yichao Yan, Junjie Li, Shengcai Liao, Jie Qin, Bingbing Ni, Xiaokang Yang
In the meantime, we design an adaptive BN layer in the domain-invariant stream, to approximate the statistics of various unseen domains.
Domain Generalization Generalizable Person Re-identification +1
4 code implementations • 1 Sep 2021 • Yichao Yan, Jinpeng Li, Jie Qin, Shengcai Liao, Xiaokang Yang
Third, by investigating the advantages of both anchor-based and anchor-free models, we further augment AlignPS with an ROI-Align head, which significantly improves the robustness of re-id features while still keeping our model highly efficient.
Ranked #4 on Person Search on PRW
1 code implementation • ICCV 2021 • Fang Zhao, Wenhao Wang, Shengcai Liao, Ling Shao
While single-view 3D reconstruction has made significant progress benefiting from deep shape representations in recent years, garment reconstruction is still not solved well due to open surfaces, diverse topologies and complex geometric details.
no code implementations • 10 Jul 2021 • Jinpeng Li, Yichao Yan, Shengcai Liao, Xiaokang Yang, Ling Shao
Transformers have demonstrated great potential in computer vision tasks.
3 code implementations • 19 Jun 2021 • Yichao Yan, Jinpeng Li, Shengcai Liao, Jie Qin, Bingbing Ni, Xiaokang Yang, Ling Shao
This paper inventively considers weakly supervised person search with only bounding box annotations.
no code implementations • 10 Jun 2021 • Hongsong Wang, Shengcai Liao, Ling Shao
Last but not least, we introduce a region feature alignment and an instance discriminator to learn domain-invariant features for object proposals.
2 code implementations • NeurIPS 2021 • Shengcai Liao, Ling Shao
In this work, we further investigate the possibility of applying Transformers for image matching and metric learning given pairs of images.
Ranked #1 on Generalizable Person Re-identification on Market-1501 (using extra training data)
no code implementations • 27 May 2021 • Haibo Jin, Jinpeng Li, Shengcai Liao, Ling Shao
To this end, we first propose a baseline model equipped with one transformer decoder as detection head.
Ranked #5 on Face Alignment on COFW
1 code implementation • CVPR 2022 • Shengcai Liao, Ling Shao
Though online hard example mining has improved the learning efficiency to some extent, the mining in mini batches after random sampling is still limited.
Ranked #2 on Generalizable Person Re-identification on Market-1501 (using extra training data)
1 code implementation • CVPR 2021 • Yichao Yan, Jinpeng Li, Jie Qin, Song Bai, Shengcai Liao, Li Liu, Fan Zhu, Ling Shao
Person search aims to simultaneously localize and identify a query person from realistic, uncropped images, which can be regarded as the unified task of pedestrian detection and person re-identification (re-id).
Ranked #10 on Person Search on CUHK-SYSU
1 code implementation • NeurIPS 2020 • Fang Zhao, Shengcai Liao, Kaihao Zhang, Ling Shao
This paper proposes a human parsing based texture transfer model via cross-view consistency learning to generate the texture of 3D human body from a single image.
1 code implementation • 24 Nov 2020 • Wenhao Wang, Shengcai Liao, Fang Zhao, Cuicui Kang, Ling Shao
In this way, human annotations are no longer required, and it is scalable to large and diverse real-world datasets.
Generalizable Person Re-identification Unsupervised Domain Adaptation
1 code implementation • 23 Jun 2020 • Yanan Wang, Shengcai Liao, Ling Shao
To address this, we propose to automatically synthesize a large-scale person re-identification dataset following a set-up similar to real surveillance but with virtual environments, and then use the synthesized person images to train a generalizable person re-identification model.
Domain Generalization Generalizable Person Re-identification +1
1 code implementation • 11 Jun 2020 • Wenhao Wang, Fang Zhao, Shengcai Liao, Ling Shao
This paper proposes a novel light-weight module, the Attentive WaveBlock (AWB), which can be integrated into the dual networks of mutual learning to enhance the complementarity and further depress noise in the pseudo-labels.
Ranked #3 on Unsupervised Domain Adaptation on Duke to MSMT
1 code implementation • CVPR 2021 • Irtiza Hasan, Shengcai Liao, Jinpeng Li, Saad Ullah Akram, Ling Shao
Furthermore, we illustrate that diverse and dense datasets, collected by crawling the web, serve to be an efficient source of pre-training for pedestrian detection.
Ranked #3 on Pedestrian Detection on CityPersons (using extra training data)
2 code implementations • 8 Mar 2020 • Haibo Jin, Shengcai Liao, Ling Shao
The proposed model is equipped with a novel detection head based on heatmap regression, which conducts score and offset predictions simultaneously on low-resolution feature maps.
Ranked #4 on Face Alignment on COFW
1 code implementation • ECCV 2020 • Shengcai Liao, Ling Shao
In this paper, beyond representation learning, we consider how to formulate person image matching directly in deep feature maps.
Ranked #3 on Generalizable Person Re-identification on Market-1501 (MSMT17-All->mAP metric, using extra training data)
Domain Generalization Generalizable Person Re-identification +1
2 code implementations • CVPR 2019 • Wei Liu, Irtiza Hasan, Shengcai Liao
Like edges, corners, blobs and other feature detectors, the proposed detector scans for feature points all over the image, for which the convolution is naturally suited.
Ranked #8 on Pedestrian Detection on Caltech (using extra training data)
no code implementations • 24 Mar 2019 • Lei Zhang, Zhiqiang Lang, Peng Wang, Wei Wei, Shengcai Liao, Ling Shao, Yanning Zhang
To address this problem, we propose a pixel-aware deep function-mixture network for SSR, which is composed of a new class of modules, termed function-mixture (FM) blocks.
no code implementations • 13 Nov 2018 • Jianqing Zhu, Huanqiang Zeng, Jingchang Huang, Shengcai Liao, Zhen Lei, Canhui Cai, Lixin Zheng
Specifically, the same basic deep learning architecture is a shortly and densely connected convolutional neural network to extract basic feature maps of an input square vehicle image in the first stage.
Ranked #3 on Vehicle Re-Identification on VehicleID Large (mAP metric)
1 code implementation • ECCV 2018 • Wei Liu, Shengcai Liao, Weidong Hu, Xuezhi Liang, Xiao Chen
However, current single-stage detectors (e. g. SSD) have not presented competitive accuracy on common pedestrian detection benchmarks.
Ranked #11 on Pedestrian Detection on Caltech (using extra training data)
no code implementations • 7 Jul 2017 • Yang Yang, Shengcai Liao, Zhen Lei, Stan Z. Li
Then, a robust image representation based on color names is obtained by concatenating the statistical descriptors in each stripe.
1 code implementation • 9 May 2017 • Haibo Jin, Xiaobo Wang, Shengcai Liao, Stan Z. Li
However, to achieve this, existing deep models prefer to adopt image pairs or triplets to form verification loss, which is inefficient and unstable since the number of training pairs or triplets grows rapidly as the number of training data grows.
no code implementations • 16 Feb 2017 • Jianqing Zhu, Huanqiang Zeng, Shengcai Liao, Zhen Lei, Canhui Cai, Lixin Zheng
In this paper, a deep hybrid similarity learning (DHSL) method for person Re-ID based on a convolution neural network (CNN) is proposed.
no code implementations • 1 Nov 2016 • Hailin Shi, Yang Yang, Xiangyu Zhu, Shengcai Liao, Zhen Lei, Wei-Shi Zheng, Stan Z. Li
From this point of view, selecting suitable positive i. e. intra-class) training samples within a local range is critical for training the CNN embedding, especially when the data has large intra-class variations.
no code implementations • 9 May 2016 • Hailin Shi, Xiangyu Zhu, Zhen Lei, Shengcai Liao, Stan Z. Li
Deep neural networks usually benefit from unsupervised pre-training, e. g. auto-encoders.
no code implementations • ICCV 2015 • Wei-Shi Zheng, Xiang Li, Tao Xiang, Shengcai Liao, Jian-Huang Lai, Shaogang Gong
We address a new partial person re-identification (re-id) problem, where only a partial observation of a person is available for matching across different non-overlapping camera views.
no code implementations • ICCV 2015 • Shengcai Liao, Stan Z. Li
We argue that the PSD constraint provides a useful regularization to smooth the solution of the metric, and hence the learned metric is more robust than without the PSD constraint.
no code implementations • 24 Nov 2015 • Hailin Shi, Xiangyu Zhu, Shengcai Liao, Zhen Lei, Yang Yang, Stan Z. Li
In this paper, we propose a novel CNN-based method to learn a discriminative metric with good robustness to the over-fitting problem in person re-identification.
15 code implementations • 28 Nov 2014 • Dong Yi, Zhen Lei, Shengcai Liao, Stan Z. Li
The current situation in the field of face recognition is that data is more important than algorithm.
no code implementations • 18 Nov 2014 • Cuicui Kang, Shengcai Liao, Yonghao He, Jian Wang, Wenjia Niu, Shiming Xiang, Chunhong Pan
A new approach to the problem has been raised which intends to match features of different modalities directly.
2 code implementations • 6 Aug 2014 • Shengcai Liao, Anil K. Jain, Stan Z. Li
First, a new image feature called Normalized Pixel Difference (NPD) is proposed.
Ranked #6 on Face Detection on PASCAL Face
no code implementations • 5 Aug 2014 • Shengcai Liao, Zhipeng Mo, Jianqing Zhu, Yang Hu, Stan Z. Li
Person re-identification is becoming a hot research for developing both machine learning algorithms and video surveillance applications.
1 code implementation • CVPR 2015 • Shengcai Liao, Yang Hu, Xiangyu Zhu, Stan Z. Li
In this paper, we propose an effective feature representation called Local Maximal Occurrence (LOMO), and a subspace and metric learning method called Cross-view Quadratic Discriminant Analysis (XQDA).
Ranked #88 on Person Re-Identification on DukeMTMC-reID
no code implementations • 5 Jun 2014 • Dong Yi, Zhen Lei, Shengcai Liao, Stan Z. Li
For NIR-VIS problem, we produce new state-of-the-art performance on the CASIA HFB and NIR-VIS 2. 0 databases.
no code implementations • CVPR 2013 • Junjie Yan, Xucong Zhang, Zhen Lei, Shengcai Liao, Stan Z. Li
The model contains resolution aware transformations to map pedestrians in different resolutions to a common space, where a shared detector is constructed to distinguish pedestrians from background.