no code implementations • 29 Apr 2022 • Juncheng Li, Hanhui Yang, Qiaosi Yi, Faming Fang, Guangwei Gao, Tieyong Zeng, Guixu Zhang
Single image denoising (SID) has achieved significant breakthroughs with the development of deep learning.
1 code implementation • 28 Apr 2022 • Guangwei Gao, Zhengxue Wang, Juncheng Li, Wenjie Li, Yi Yu, Tieyong Zeng
Single-image super-resolution (SISR) has achieved significant breakthroughs with the development of deep learning.
no code implementations • 20 Apr 2022 • Longguang Wang, Yulan Guo, Yingqian Wang, Juncheng Li, Shuhang Gu, Radu Timofte
In this paper, we summarize the 1st NTIRE challenge on stereo image super-resolution (restoration of rich details in a pair of low-resolution stereo images) with a focus on new solutions and results.
no code implementations • 19 Apr 2022 • Guangwei Gao, Zixiang Xu, Juncheng Li, Jian Yang, Tieyong Zeng, Guo-Jun Qi
Specifically, we first devise a novel Local-Global Feature Cooperation Module (LGCM), which is composed of a Facial Structure Attention Unit (FSAU) and a Transformer block, to promote the consistency of local facial detail and global facial structure restoration simultaneously.
1 code implementation • 24 Mar 2022 • Juncheng Li, Junlin Xie, Long Qian, Linchao Zhu, Siliang Tang, Fei Wu, Yi Yang, Yueting Zhuang, Xin Eric Wang
To systematically measure the compositional generalizability of temporal grounding models, we introduce a new Compositional Temporal Grounding task and construct two new dataset splits, i. e., Charades-CG and ActivityNet-CG.
1 code implementation • 16 Dec 2021 • Guangwei Gao, Wenjie Li, Juncheng Li, Fei Wu, Huimin Lu, Yi Yu
Convolutional neural networks based single-image super-resolution (SISR) has made great progress in recent years.
no code implementations • 13 Dec 2021 • Wenqiao Zhang, Haochen Shi, Jiannan Guo, Shengyu Zhang, Qingpeng Cai, Juncheng Li, Sihui Luo, Yueting Zhuang
We propose the Multimodal relAtional Graph adversarIal inferenCe (MAGIC) framework for diverse and unpaired TextCap.
1 code implementation • 29 Sep 2021 • Juncheng Li, Zehua Pei, Tieyong Zeng
In this survey, we give an overview of DL-based SISR methods and group them according to their targets, such as reconstruction efficiency, reconstruction accuracy, and perceptual accuracy.
1 code implementation • 2 Sep 2021 • Guangwei Gao, Guoan Xu, Juncheng Li, Yi Yu, Huimin Lu, Jian Yang
Specifically, FBSNet employs a symmetrical encoder-decoder structure with two branches, semantic information branch and spatial detail branch.
1 code implementation • 25 Aug 2021 • Zhisheng Lu, Juncheng Li, Hong Liu, Chaoyan Huang, Linlin Zhang, Tieyong Zeng
LTB is composed of a series of Efficient Transformers (ET), which occupies a small GPU memory occupation, thanks to the specially designed Efficient Multi-Head Attention (EMHA).
1 code implementation • ICCV 2021 • Qiaosi Yi, Juncheng Li, Qinyan Dai, Faming Fang, Guixu Zhang, Tieyong Zeng
Although these methods can remove part of the rain streaks, it is difficult for them to adapt to real-world scenarios and restore high-quality rain-free images with clear and accurate structures.
no code implementations • ICCV 2021 • Juncheng Li, Siliang Tang, Linchao Zhu, Haochen Shi, Xuanwen Huang, Fei Wu, Yi Yang, Yueting Zhuang
Secondly, we introduce semantic coherence learning to explicitly encourage the semantic coherence of the adaptive hierarchical graph network from three hierarchies.
1 code implementation • 2 Jun 2021 • Qinyan Dai, Juncheng Li, Qiaosi Yi, Faming Fang, Guixu Zhang
Besides the cross-view information exploitation in the low-resolution (LR) space, HR representations produced by the SR process are utilized to perform HR disparity estimation with higher accuracy, through which the HR features can be aggregated to generate a finer SR result.
no code implementations • 24 Mar 2021 • Zhengxue Wang, Guangwei Gao, Juncheng Li, Yi Yu, Huimin Lu
Recently, the single image super-resolution (SISR) approaches with deep and complex convolutional neural network structures have achieved promising performance.
no code implementations • 24 Feb 2021 • Qiaosi Yi, Juncheng Li, Faming Fang, Aiwen Jiang, Guixu Zhang
To achieve this, we propose a Multi-scale Topological Network (MSTN) to fully explore the features at different scales.
no code implementations • 5 Jan 2021 • Qiaosi Yi, Yunxing Liu, Aiwen Jiang, Juncheng Li, Kangfu Mei, Mingwen Wang
Although the emergence of deep learning has greatly promoted the development of this field, crowd counting under cluttered background is still a serious challenge.
no code implementations • 1 Jan 2021 • Dong Chen, Lingfei Wu, Siliang Tang, Fangli Xu, Juncheng Li, Chang Zong, Chilie Tan, Yueting Zhuang
In particular, we first cast the meta-overfitting problem (overfitting on sampling and label noise) as a gradient noise problem since few available samples cause meta-learner to overfit on existing examples (clean or corrupted) of an individual task at every gradient step.
no code implementations • 12 Sep 2020 • Ze Cheng, Juncheng Li, Chenxu Wang, Jixuan Gu, Hao Xu, Xinjian Li, Florian Metze
In this paper, we provide a theoretical explanation that low total correlation of sampled representation cannot guarantee low total correlation of the mean representation.
1 code implementation • 30 Aug 2020 • Juncheng Li, Faming Fang, Jiaqian Li, Kangfu Mei, Guixu Zhang
Among them, MDCB aims to detect multi-scale features and maximize the use of image features flow at different scales, HFDB focuses on adaptively recalibrate channel-wise feature responses to achieve feature distillation, and DRB attempts to reconstruct SR images with different upsampling factors in a single model.
no code implementations • 11 Aug 2020 • Jiacheng Li, Siliang Tang, Juncheng Li, Jun Xiao, Fei Wu, ShiLiang Pu, Yueting Zhuang
In this paper, we focus on enhancing the generalization ability of the VIST model by considering the few-shot setting.
no code implementations • 24 Jun 2020 • Kangfu Mei, Yao Lu, Qiaosi Yi, Hao-Yu Wu, Juncheng Li, Rui Huang
Perceptual learning approaches like perceptual loss are empirically powerful for such tasks but they usually rely on the pre-trained classification network to provide features, which are not necessarily optimal in terms of visual perception of image transformation.
1 code implementation • 26 Feb 2020 • Xinjian Li, Siddharth Dalmia, Juncheng Li, Matthew Lee, Patrick Littell, Jiali Yao, Antonios Anastasopoulos, David R. Mortensen, Graham Neubig, Alan W. black, Florian Metze
Multilingual models can improve language processing, particularly for low resource situations, by sharing parameters across languages.
no code implementations • 26 Feb 2020 • Xinjian Li, Siddharth Dalmia, David R. Mortensen, Juncheng Li, Alan W. black, Florian Metze
The difficulty of this task is that phoneme inventories often differ between the training languages and the target language, making it infeasible to recognize unseen phonemes.
no code implementations • 25 Feb 2020 • Han Wang, Juncheng Li, Maopeng Ran, Lihua Xie
Our method is compared with the state-of-the-art loop closure detection methods and the results show that it outperforms the traditional methods at both recall rate and speed.
1 code implementation • 19 Nov 2019 • Kangfu Mei, Juncheng Li, Jiajie Zhang, Hao-Yu Wu, Jie Li, Rui Huang
However, plenty of studies have shown that global information is crucial for image restoration tasks like image demosaicing and enhancing.
no code implementations • CVPR 2020 • Juncheng Li, Xin Wang, Siliang Tang, Haizhou Shi, Fei Wu, Yueting Zhuang, William Yang Wang
Visual navigation is a task of training an embodied agent by intelligently navigating to a target object (e. g., television) using only visual observations.
reinforcement-learning
Unsupervised Reinforcement Learning
+1
no code implementations • 5 Aug 2019 • Juncheng Li, Siliang Tang, Fei Wu, Yueting Zhuang
The experimental results and further analysis prove that the agent with the MIND module is superior to its counterparts not only in EQA performance but in many other aspects such as route planning, behavioral interpretation, and the ability to generalize from a few examples.
1 code implementation • 21 Mar 2019 • Juncheng Li, Frank R. Schmidt, J. Zico Kolter
In this work, we consider an alternative question: is it possible to fool deep classifiers, over all perceived objects of a certain type, by physically manipulating the camera itself?
3 code implementations • 22 Oct 2018 • Yun Wang, Juncheng Li, Florian Metze
This paper compares five types of pooling functions both theoretically and experimentally, with special focus on their performance of localization.
Sound Audio and Speech Processing
1 code implementation • 4 Oct 2018 • Kangfu Mei, Aiwen Jiang, Juncheng Li, Mingwen Wang
Most of them follow a classic atmospheric scattering model which is an elegant simplified physical model based on the assumption of single-scattering and homogeneous atmospheric medium.
1 code implementation • 3 Oct 2018 • Kangfu Mei, Aiwen Jiang, Juncheng Li, Jihua Ye, Mingwen Wang
Recent works on single-image super-resolution are concentrated on improving performance through enhancing spatial encoding between convolutional layers.
1 code implementation • ECCV 2018 • Juncheng Li, Faming Fang, Kangfu Mei, Guixu Zhang
Meanwhile, we let these features interact with each other to get the most efficacious image information, we call this structure Multi-scale Residual Block (MSRB).
1 code implementation • ICMR 2018 • Niluthpol Chowdhury Mithun, Juncheng Li, Florian Metze, Amit K. Roy-Chowdhury
Constructing a joint representation invariant across different modalities (e. g., video, language) is of significant importance in many multimedia applications.
Ranked #15 on
Video Retrieval
on MSR-VTT
1 code implementation • 20 Mar 2017 • Juncheng Li, Wei Dai, Florian Metze, Shuhui Qu, Samarjit Das
On these features, we apply five models: Gaussian Mixture Model (GMM), Deep Neural Network (DNN), Recurrent Neural Network (RNN), Convolutional Deep Neural Net- work (CNN) and i-vector.
no code implementations • 29 Nov 2016 • Shuhui Qu, Juncheng Li, Wei Dai, Samarjit Das
Based on the procedure of log Mel-filter banks, we design a filter bank learning layer.
8 code implementations • 1 Oct 2016 • Wei Dai, Chia Dai, Shuhui Qu, Juncheng Li, Samarjit Das
Our CNNs, with up to 34 weight layers, are efficient to optimize over very long sequences (e. g., vector of size 32000), necessary for processing acoustic waveforms.