Paper Detail

Lip Forcing: Few-Step Autoregressive Diffusion for Real-time Lip Synchronization

Paul Hyunbin Cho, Jinhyuk Jang, SeokYoung Lee, Joungbin Lee, Siyoon Jin, Heeseong Shin, Jung Yi, Yunjin Park, Chulmin Park, Seungryong Kim

arxiv Score 4.3

Published 2026-06-09 · First seen 2026-06-10

General AI

Abstract

Diffusion-based lip synchronization models achieve strong visual quality and audio-visual alignment, but full-sequence bidirectional attention and many denoising steps make them impractical for real-time inference. We present Lip Forcing, to our knowledge the first autoregressive diffusion method for video-to-video (V2V) lip synchronization, which distills a 14B audio-conditioned bidirectional video diffusion teacher into causal students. At inference, the students generate each chunk in only two denoising steps without inference-time CFG, enabling real-time lip synchronization. A lip-sync-specific teacher-trajectory analysis reveals a CFG fidelity-sync tradeoff: no-CFG predictions favor reference fidelity, whereas CFG-guided predictions favor synchronization within a mid-trajectory band. Lip Forcing translates this finding into three analysis-derived components: Sync-Window DMD, a two-step inference schedule, and a SyncNet-based reward. We validate Lip Forcing at two student scales, both distilled from the 14B teacher. The 1.3B student crosses into real-time streaming at 31 FPS, $17.6\times$ faster than its same-scale bidirectional model. The 14B student, the largest diffusion model reported for V2V lip synchronization, runs $39.8\times$ faster than its teacher at comparable reference fidelity. Time-to-first-frame is sub-millisecond at both scales, far below every diffusion baseline.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
later
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{cho2026lip,
  title = {Lip Forcing: Few-Step Autoregressive Diffusion for Real-time Lip Synchronization},
  author = {Paul Hyunbin Cho and Jinhyuk Jang and SeokYoung Lee and Joungbin Lee and Siyoon Jin and Heeseong Shin and Jung Yi and Yunjin Park and Chulmin Park and Seungryong Kim},
  year = {2026},
  abstract = {Diffusion-based lip synchronization models achieve strong visual quality and audio-visual alignment, but full-sequence bidirectional attention and many denoising steps make them impractical for real-time inference. We present Lip Forcing, to our knowledge the first autoregressive diffusion method for video-to-video (V2V) lip synchronization, which distills a 14B audio-conditioned bidirectional video diffusion teacher into causal students. At inference, the students generate each chunk in only tw},
  url = {https://arxiv.org/abs/2606.11180},
  keywords = {cs.CV, diffusion models, lip synchronization, video-to-video, bidirectional attention, denoising steps, causal students, teacher-student distillation, inference-time CFG, SyncNet, time-to-first-frame, code available, huggingface daily},
  eprint = {2606.11180},
  archiveprefix = {arXiv},
}

Metadata

{}