Search Results for author: Fei Du

Found 6 papers, 4 papers with code

SwinRDM: Integrate SwinRNN with Diffusion Model towards High-Resolution and High-Quality Weather Forecasting

no code implementations5 Jun 2023 Lei Chen, Fei Du, Yuan Hu, Fan Wang, Zhibin Wang

Recurrent predictions for future atmospheric fields are firstly performed at 1. 40625-degree resolution, and then a diffusion-based super-resolution model is leveraged to recover the high spatial resolution and finer-scale atmospheric details.

Super-Resolution Weather Forecasting

Global and Local Mixture Consistency Cumulative Learning for Long-tailed Visual Recognitions

1 code implementation CVPR 2023 Fei Du, Peng Yang, Qi Jia, Fengtao Nan, Xiaoting Chen, Yun Yang

In this paper, our goal is to design a simple learning paradigm for long-tail visual recognition, which not only improves the robustness of the feature extractor but also alleviates the bias of the classifier towards head classes while reducing the training skills and overhead.

Long-tail Learning

FAKD: Feature Augmented Knowledge Distillation for Semantic Segmentation

1 code implementation30 Aug 2022 Jianlong Yuan, Qian Qi, Fei Du, Zhibin Wang, Fan Wang, Yifan Liu

Inspired by the recent progress on semantic directions on feature-space, we propose to include augmentations in feature space for efficient distillation.

Knowledge Distillation Semantic Segmentation

1st Place Solution to ECCV-TAO-2020: Detect and Represent Any Object for Tracking

1 code implementation20 Jan 2021 Fei Du, Bo Xu, Jiasheng Tang, Yuqi Zhang, Fan Wang, Hao Li

We extend the classical tracking-by-detection paradigm to this tracking-any-object task.

 Ranked #1 on Multi-Object Tracking on TAO (using extra training data)

Multi-Object Tracking

Correlation-Guided Attention for Corner Detection Based Visual Tracking

no code implementations CVPR 2020 Fei Du, Peng Liu, Wei Zhao, Xianglong Tang

Accurate bounding box estimation has recently attracted much attention in the tracking community because traditional multi-scale search strategies cannot estimate tight bounding boxes in many challenging scenarios involving changes to the target.

Visual Tracking

Cannot find the paper you are looking for? You can Submit a new open access paper.