Qing Guo

exposure-fusion-shadow-removal. We propose a new method for effective shadow removal by regarding it as an exposure fusion problem.

efficientderain. We propose EfficientDerain for high-efficiency single-image deraining at AAAI 2021. The latest extended version has been accepted by IJCV at 2024. Please check the released code at https://github.com/tsingqguo/efficientderainplus and paper at https://arxiv.org/pdf/2201.02366.

misf. Python

DSiam. Learning Dynamic Siamese Network for Visual Object Tracking

inpaint4shadow. We propose the shadow-guided inpainting task to take advantage of the shadow removal and image inpainting.

ABBA. Python

bgmix. We propose a novel data augmentation by enriching the backgrounds for change detection in a weakly-superivsed way.

AttackTracker. Python

scenetap. Python

jpgnet. We proposed a novel framework for image inpainting. https://arxiv.org/abs/2107.04281

jadena. Official implementation of "Can You Spot the Chameleon? Adversarially Camouflaging Images from Co-Salient Object Detection" in CVPR 2022.

ABA. We propose the adversarial blur attack (ABA) against visual object tracking.

irad. We introduce a novel approach to counter adversarial attacks, namely, image resampling. The underlying rationale behind our idea is that image resampling can alleviate the influence of adversarial perturbations while preserving essential semantic information, thereby conferring an inherent advantage in defending against adversarial attacks.

coljailbreak. We propose ColJailBreak, which serves as an innovative jailbreaking framework designed to bypass65 safety filters in commercial T2I models by initially generating safe content and then injecting66 unsafe elements through editing.

robustOT. We propose a novel adversarial tracking defense method throug spatial-temporal implicit representation, which can enhance the adversarial robustness of tracking methods under different tracking attacks. The work has been published in ICLR 2024.

efficientderainplus. We further extend the efficientderain in https://github.com/tsingqguo/efficientderain via a novel predictive filtering framework. This work has been accepted by IJCV at 2024.

deeprhythm. Python

sair. We propose the semantic-aware implicit representation by learning semantic-aware implicit representation (SAIR), that is, we make the implicit representation of each pixel rely on both its appearance and semantic information (e.g., which object does the pixel belong to). This work is publised in ECCV 2024.

anglerocl. We introduce Angle-Robust Concept Learning (AngleRoCL), a simple and flexible approach that learns a generalizable concept representing the capability of generating angle-robust patches. The learned concept can be incorporated into textual prompts and guides T2I to generate patches with their attack inherently resistant to viewpoint variations.

sharel. We propose a shadow-removal benchmark dataset (i.e., SHAREL) to explore the mutual influence of shadow removal and facial landmark detection tasks.

evadingfakedetector. We propose a statistical consistency attack (StatAttack) against diverse DeepFake detectors.

augmix. AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty

resample4defense. We have identified a novel adversarial defense solution, i.e., image resampling, which can break the adversarial textures while maintaining the main semantic information in the input image. This work has been accepted to ICLR 2024.

ccotssr. abc

frequency-tuned-ACM. This repository contains the demos of frequency-tuned ACM.

a-PyTorch-Tutorial-to-Object-Detection. SSD: Single Shot MultiBox Detector | a PyTorch Tutorial to Object Detection

MSiam. Try to build a new tracking framework with Siamese network

tsingqguo.github.io. Homepage of Qing Guo