BCMI
Change the repository type filter
All
Repositories list
75 repositories
- [CVPR 2024] The dataset, code, and model for our paper "Shadow Generation for Composite Image Using Diffusion Model", CVPR, 2024.
Light-A-Video
Public[ICCV 2025] Light-A-Video: Training-free Video Relighting via Progressive Light FusionCausal-VidQA
Public[CVPR 2022] A large-scale public benchmark dataset for video question-answering, especially about evidence and commonsense reasoning. The code used in our paper "From Representation to Reasoning: Towards both Evidence and Commonsense Reasoning for Video Question-Answering", CVPR2022.- A simple baseline for image composition using text-guided inpainting model
- A controllable image composition model which could be used for image blending, image harmonization, view synthesis.
- A curated list of papers, code, and resources pertaining to generative image composition or object insertion.
- A curated list of papers, code, and resources pertaining to object shadow generation.
- [ICME2021]The first work on Deep Inharmonious Region Localization, which can help image harmonization in an adversarial way.
- [AAAI 2022] MadisNet: Inharmonious Region Localization by Magnifying Domain Discrepancy
- [AAAI 2022] The first dataset on foreground object shadow generation for image composition in real-world scenes. The code used in our paper "Shadow Generation for Composite Image in Real-world Scenes", AAAI2022. Useful for shadow generation, shadow removal, image composition, etc.
- A virtual accessory try-on dataset which could be used for image composition
- [IJCAI 2022] The first public benchmark dataset for video harmonization. The code used in our paper "Deep Video Harmonization with Color Mapping Consistency", IJCAI 2022.
- An unofficial implementation of the paper "ObjectStitch: Object Compositing with Diffusion Model", CVPR 2023.
- A discriminative object placement approach
- [ICCV 2023] The color checker based harmonization dataset contributed in our paper "Deep Image Harmonization with Globally Guided Feature Transformation and Relation Distillation", ICCV2023.
TopNet-Object-Placement
Public- [AAAI 2023] Painterly image harmonization in both spatial domain and frequency domain.
- [CVPR 2022] We unify pixel-to-pixel transformation and color-to-color transformation in a coherent framework for high-resolution image harmonization. We also release 100 high-resolution real composite images for evaluation.
- [ICCV 2023] The code used in our paper "Deep Image Harmonization with Learnable Augmentation", ICCV2023.
- [ICCV 2023] The datasets and code used in our paper "Foreground Object Search by Distilling Composite Image Feature", ICCV2023.
- BargainNet: Background-Guided Domain Translation for Image Harmonization. Useful for Image harmonization, image composition, etc.
- The first rendered image harmonization dataset. Used in our paper "CharmNet: Deep Image Harmonization by Bridging the Reality Gap". Useful for Image harmonization, image composition, etc.
- [AAAI 2024] The dataset used in our paper "Shadow Generation with Decomposed Mask Prediction and Attentive Shadow Filling", AAAI 2024.
- The first dataset of composite images with rationality score indicating whether the object placement in a composite image is reasonable.