FaithfulFaces: Pose-Faithful Facial Identity Preservation for Text-to-Video Generation

Hugging Face Daily Papers Papers
text-to-video facial-identity pose-preservation computer-vision deep-learning

Summary

FaithfulFaces is a new framework for text-to-video generation that preserves facial identity consistency across varying poses and occlusions using pose-shared alignment and Euler angle embeddings.

Identity-preserving text-to-video generation (IPT2V) empowers users to produce diverse and imaginative videos with consistent human facial identity. Despite recent progress, existing methods often suffer from significant identity distortion under large facial pose variations or facial occlusions. In this paper, we propose FaithfulFaces, a pose-faithful facial identity preservation learning framework to improve IPT2V in complex dynamic scenes. The key of FaithfulFaces is a pose-shared identity aligner that refines and aligns facial poses across distinct views via a pose-shared dictionary and a pose variation-identity invariance constraint. By mapping single-view inputs into a global facial pose representation with explicit Euler angle embeddings, FaithfulFaces provides a pose-faithful facial prior that guides generative foundations toward robust identity-preserving generation. In particular, we develop a specialized pipeline to curate a high-quality video dataset featuring substantial facial pose diversity. Extensive experiments demonstrate that FaithfulFaces achieves state-of-the-art performance, maintaining superior identity consistency and structural clarity even as pose changes and occlusions occur.
Original Article
View Cached Full Text

Cached at: 05/13/26, 12:15 PM

Paper page - FaithfulFaces: Pose-Faithful Facial Identity Preservation for Text-to-Video Generation

Source: https://huggingface.co/papers/2605.04702

Abstract

FaithfulFaces is a pose-faithful facial identity preservation framework that improves identity consistency in text-to-video generation through pose-shared alignment and explicit Euler angle embeddings.

Identity-preservingtext-to-video generation(IPT2V) empowers users to produce diverse and imaginative videos with consistent human facial identity. Despite recent progress, existing methods often suffer from significant identity distortion under large facial pose variations or facial occlusions. In this paper, we propose FaithfulFaces, apose-faithful facial identity preservationlearning framework to improve IPT2V in complex dynamic scenes. The key of FaithfulFaces is apose-shared identity alignerthat refines and aligns facial poses across distinct views via apose-shared dictionaryand apose variation-identity invariance constraint. By mapping single-view inputs into aglobal facial pose representationwith explicitEuler angle embeddings, FaithfulFaces provides a pose-faithful facial prior that guides generative foundations toward robust identity-preserving generation. In particular, we develop a specialized pipeline to curate a high-quality video dataset featuring substantial facial pose diversity. Extensive experiments demonstrate that FaithfulFaces achieves state-of-the-art performance, maintaining superioridentity consistencyandstructural clarityeven as pose changes and occlusions occur.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2605\.04702

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.04702 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.04702 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.04702 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles