Cross-Domain Few-Shot Object Detection via Enhanced Open-Set Object Detector

no code implementations • 5 Feb 2024 • Yuqian Fu, Yu Wang, Yixuan Pan, Lian Huai, Xingyu Qiu, Zeyu Shangguan, Tong Liu, Yanwei Fu, Luc van Gool, Xingqun Jiang

This paper studies the challenging cross-domain few-shot object detection (CD-FSOD), aiming to develop an accurate object detector for novel domains with minimal labeled examples.

Cross-Domain Few-Shot Few-Shot Object Detection +3

Decoupled DETR For Few-shot Object Detection

no code implementations • 20 Nov 2023 • Zeyu Shangguan, Lian Huai, Tong Liu, Xingqun Jiang

We also explore various types of skip connection between the encoder and decoder for DETR.

Few-Shot Object Detection Meta-Learning +2

Few-shot Object Detection with Refined Contrastive Learning

no code implementations • 24 Nov 2022 • Zeyu Shangguan, Lian Huai, Tong Liu, Xingqun Jiang

A pre-determination component is introduced to find out the Resemblance Group from novel classes which contains confusable classes.

Contrastive Learning Few-Shot Object Detection +2

Distinctive Self-Similar Object Detection

no code implementations • 20 Nov 2022 • Zeyu Shangguan, Bocheng Hu, Guohua Dai, Yuyu Liu, Darun Tang, Xingqun Jiang

However, objects such as fire and smoke, pose challenges to object detection because of their non-solid and various shapes, and consequently difficult to truly meet requirements in practical fire prevention and control.

Object object-detection +1

edge-SR: Super-Resolution For The Masses

1 code implementation • 23 Aug 2021 • Pablo Navarrete Michelini, Yunhua Lu, Xingqun Jiang

We explore possible solutions to this problem with the aim to fill the gap between classic upscalers and small deep learning configurations.

Image Super-Resolution

Back-Projection Pipeline

no code implementations • 25 Jan 2021 • Pablo Navarrete Michelini, Hanwen Liu, Yunhua Lu, Xingqun Jiang

We propose a simple extension of residual networks that works simultaneously in multiple resolutions.

Rain Removal Super-Resolution

Multi-Grid Back-Projection Networks

no code implementations • 1 Jan 2021 • Pablo Navarrete Michelini, Wenbin Chen, Hanwen Liu, Dan Zhu, Xingqun Jiang

For this target we propose a strategy using noise inputs in different resolution scales to control the amount of artificial details generated in the output.

A Tour of Convolutional Networks Guided by Linear Interpreters

1 code implementation • ICCV 2019 • Pablo Navarrete Michelini, Hanwen Liu, Yunhua Lu, Xingqun Jiang

These units are the "articulations" that allow the network to adapt to the input.

Image-to-Image Translation Super-Resolution