metadata
language:
- en
license: apache-2.0
tags:
- image-to-video
- video-generation
- identity-preservation
- view-consistency
- diffusion
- consid-gen
pipeline_tag: image-to-video
library_name: diffsynth
ConsID-Gen
ConsID-Gen: View-Consistent and Identity-Preserving Image-to-Video Generation
Mingyang Wu, Ashirbad Mishra, Soumik Dey, Shuo Xing, Naveen Ravipati, Hansi Wu, Binbin Li, Zhengzhong Tu (2026)
Accepted by CVPR 2026.
Summary
This repository contains the model checkpoint for our paper:
ConsID-Gen: View-Consistent and Identity-Preserving Image-to-Video Generation.
ConsID-Gen focuses on generating videos that maintain:
- strong identity preservation,
- cross-view consistency,
- temporal coherence.
Files
model.safetensors: Main model checkpoint.
Usage
Please refer to the project scripts for training/inference entry points (for example run_train_considgen.py and run_inference_considgen.py) and adapt paths/configs to your environment.
Citation
@misc{wu2026considgenviewconsistentidentitypreservingimagetovideo,
title={ConsID-Gen: View-Consistent and Identity-Preserving Image-to-Video Generation},
author={Mingyang Wu and Ashirbad Mishra and Soumik Dey and Shuo Xing and Naveen Ravipati and Hansi Wu and Binbin Li and Zhengzhong Tu},
year={2026},
eprint={2602.10113},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2602.10113},
}