# HarmoView: Harmonizing Multi-View Constraints for Identity-Consistent Video Generation

arXiv: https://arxiv.org/abs/2606.10839 · HTML: https://mlsanity.com/paper/2606.10839/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 25 references that are in the map.
- First posted on arXiv 9 Jun 2026.

## Builds on (most cited first, 20 of 20)

- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space](https://mlsanity.com/paper/2506.15742/) (2025)
- [VACE: All-in-One Video Creation and Editing](https://mlsanity.com/paper/2503.07598/) (2025)
- [NIMA: Neural Image Assessment](https://mlsanity.com/paper/1709.05424/) (2017)
- [OmniHuman-1: Rethinking the Scaling-Up of One-Stage Conditioned Human Animation Models](https://mlsanity.com/paper/2502.01061/) (2025)
- [Phantom: Subject-consistent video generation via cross-modal alignment](https://mlsanity.com/paper/2502.11079/) (2025)
- [ID-Animator: Zero-Shot Identity-Preserving Human Video Generation](https://mlsanity.com/paper/2404.15275/) (2024)
- [Hallo3: Highly Dynamic and Realistic Portrait Image Animation with Video Diffusion Transformer](https://mlsanity.com/paper/2412.00733/) (2024)
- [HunyuanCustom: A Multimodal-Driven Architecture for Customized Video Generation](https://mlsanity.com/paper/2505.04512/) (2025)
- [FantasyTalking: Realistic Talking Portrait Generation via Coherent Motion Synthesis](https://mlsanity.com/paper/2504.04842/) (2025)
- [OpenS2V-Nexus: A Detailed Benchmark and Million-Scale Dataset for Subject-to-Video Generation](https://mlsanity.com/paper/2505.20292/) (2025)
- [HuMo: Human-Centric Video Generation via Collaborative Multi-Modal Conditioning](https://mlsanity.com/paper/2509.08519/) (2025)
- [MeGA: Hybrid Mesh-Gaussian Head Avatar for High-Fidelity Rendering and Head Editing](https://mlsanity.com/paper/2404.19026/) (2024)
- [MVGenMaster: Scaling Multi-View Generation from Any Image via 3D Priors Enhanced Diffusion Model](https://mlsanity.com/paper/2411.16157/) (2024)
- [Hallo4: High-Fidelity Dynamic Portrait Animation via Direct Preference Optimization](https://mlsanity.com/paper/2505.23525/) (2025)
- [Kaleido: Open-Sourced Multi-Subject Reference Video Generation Model](https://mlsanity.com/paper/2510.18573/) (2025)
- [RefAlign: Representation Alignment for Reference-to-Video Generation](https://mlsanity.com/paper/2603.25743/) (2026)
- [MV-S2V: Multi-View Subject-Consistent Video Generation](https://mlsanity.com/paper/2601.17756/) (2026)
- [CausNVS: Autoregressive Multi-view Diffusion for Flexible 3D Novel View Synthesis](https://mlsanity.com/paper/2509.06579/) (2025)
- [Bootstrap3D: Improving Multi-view Diffusion Model with Synthetic Data](https://mlsanity.com/paper/2406.00093/) (2024)

## Built on by (most cited first, 1 of 1)

- [OmniVBench: A Benchmark and Large-Scale Dataset for Omni Reference-to-Video Generation](https://mlsanity.com/paper/2609.22069/) (2026)

## Abstract (arXiv)

Current identity-consistent video generation methods struggle to preserve appearance fidelity under large viewpoint changes. While introducing multi-view reference input offers a natural solution, progress remains constrained by the lack of effective frameworks for multi-view inputs and the scarcity of multi-view data. We address these challenges by proposing HarmoView, a robust framework for identity-consistent video generation that effectively integrates multi-view cues through three architectural refinements complemented by a staged training curriculum. Specifically, we first introduce Multi-level Feature Injection to anchor identity fidelity; by injecting raw ViT features from frontal references alongside text tokens via cross-attention, MFI provides persistent low-level appearance anchors that complement the high-level identity features within DiT blocks, leading to enhanced identity preservation. Then, we employ learnable proxy tokens to unify heterogeneous reference layouts across single-/multi-view settings while simultaneously resolving the reference-view mismatch problem. Jump-RoPE is further developed for identity-wise feature isolation to reduce identity crosstalk. To activate these structural capabilities while preserving the original generative priors, we propose the Progressive View Curriculum. This four-stage training strategy employs view dropout to facilitate a stable transition from vanilla T2V generation to high-fidelity, identity-persistent spatial reasoning. Furthermore, we construct a large-scale multi-view dataset to address the issue of data scarcity. Extensive evaluation on our multi-view benchmark, comprising 100 manually-curated cases spanning 52 unique identities, demonstrates that HarmoView significantly outperforms open-source baselines and matches leading closed-source engines, achieving state-of-the-art performance in identity-consistent video generation.