arXiv Computer Vision By Kaixing Yang, Xulong Tang, Ronghui Li, Xiangyue Zhang, Ziqiao Peng, Mingshuang Luo, Chenxi Zhao, Haoyu Wu, Ju Fan, Hongyan Liu, Puwei Wang, Zijiao Zeng, Jun He

FlashMotion: Ultra-Fast Long-Term 3D Motion Generation

Read the original on arXiv Computer Vision →

FlashMotion is a new framework for generating high‑quality, long‑horizon 3D human motion quickly enough for real‑time interactive use. It combines a 3D motion‑space MeanFlow formulation with adaptive soft masking for efficient few‑step generation, and a lightweight non‑autoregressive BiMamba architecture that improves per‑step speed while maintaining cross‑modal alignment. Experiments show state‑of‑the‑art performance on text‑to‑motion, music‑to‑dance, and speech‑to‑gesture tasks, with code released publicly.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

arXiv Computer Vision
Aug 27

InteractGesture: Progressive Chunk Guidance for Continuous Streaming Co-Speech Gesture Control

InteractGesture is a model‑agnostic, inference‑time method that enables fine‑grained spatial control of individual joints in continuous streaming co‑speech gesture generation. It guides diffusion sampler latent estimates through a differentiable RVQ‑VAE decoder, backpropagating spatial control gradients to adjust motion latents during sampling. To address chunk‑wise dependency issues in streaming generation, the method introduces Progressive Chunk Guidance, a chunk‑window strategy that keeps an active set of editable chunk latents with staggered delays, allowing spatial constraints to propagate gradients backward across chunk boundaries and reducing boundary inconsistencies.

By Ekkasit Pinyoanuntapong, Ajinkya Deogade, Paul Streli, Wenjing Zhang, Joanna Materzynska, Pu Wang, Vittorio Ferrari, Jie Shen
arXiv Computer Vision
Sep 11

Multi-Modal Controlled Coherent Motion Generation

The paper introduces MOCO, a diffusion-based framework that generates 3D avatar motions from concurrent multimodal inputs such as speech audio, text descriptions, and trajectory data. MOCO decouples motion generation by independently producing modality-specific motions at each denoising step and then assembling them according to spatial rules, iteratively refining the combined motion. This approach yields coherent, lifelike, and synchronized movements, outperforming existing baselines on a multimodal benchmark.

By Yifei Liu, Qiong Cao, Hongwei Yi, Huaiguang Jiang, Changxing Ding
Hugging Face Trending Papers
Aug 10

UniMoFlow: Grounding Instruction-Driven 3D Human Motion Editing in Generation

Instruction-driven editing of 3D human motion requires precise spatiotemporal localization, rich semantic grounding, and strict preservation of unmodified content. Existing methods either resort to training-free adaptation of generative models or rely solely on triplet supervision; however, adaptation often yields suboptimal control, and manually curated triplet datasets remain severely limited in scale and semantic diversity.