AI Paper Atlas

Research map / Video / Video generation

Soul: Breathe Life into Digital Human for High-fidelity Long-term Multimodal Animation

Jiangning Zhang, Junwei Zhu, Zhenye Gan, Donghao Luo, Chuming Lin, FeiFan Xu, Xu Peng, Jianlong Hu, Yuansen Liu, Yijia Hong, Weijian Cao, Han Feng and 5 more

CVPR 2026 · Video › Video generation · approach: Diffusion-based human image animation · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train on Soul-1M and use Soul design instead of prior open-source avatar methodsSoul · Soul-BenchIdentity Consistence: 0.750 -> 0.763Improved
use Soul instead of Wan-S2VSoul · Soul-BenchLSE-D: 5.455 -> 0.130Improved
use Soul instead of InfiniteTalkSoul · Soul-BenchLSE-C: 8.48 -> 6.82Hurt
use Soul instead of StableAvatarSoul · Soul-BenchVideo Quality: 71.40 -> 72.60Improved
add step and CFG distillation (ablation)Soul · Soul-BenchFull-model latency (s): 1019.2s -> 135.3sImproved
add step and CFG distillation (ablation)Soul · Soul-BenchIdentity Consistence: 0.763 -> 0.696Hurt
add step and CFG distillation (ablation)Soul · Soul-BenchVideo Quality: 72.60 -> 71.90Hurt
add efficient eVAE decoder after distillation (ablation)Soul · Soul-BenchFull-model latency (s): 135.3s -> 89.4sImproved
add efficient eVAE decoder after distillation (ablation)Soul · Soul-BenchIdentity Consistence: 0.696 -> 0.702No clear effect
replace Wan2.2-5B VAE decoder with eVAE-35M (ablation)Wan2.2-5B VAE · 720x1280 reconstructionLPIPS: 0.0324 -> 0.052Hurt
replace Wan2.2-5B VAE decoder with eVAE-35M (ablation)Wan2.2-5B VAE · 720x1280 reconstructionDecoder speedup: 1.0x -> 8.1xImproved
add threshold-aware codebook replacement (ablation)Soul long-term generation · long video generation, qualitativevisual quality over timeImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy