11 papers
Stabilizing Instruction Supervision for Instruct-TTS via Controllable Diversification and Drift Filtering
Yizhong Geng, Kecan Mao, Qifei Li +6
Instruct-TTS systems expand structured style labels into natural-language training instructions through LLM rewriting, yet we find that over 40% of unconstrained rewrites contain s…
CLASVS: Continuous-Latent Autoregression for Melody-Preserving Lyric Editing in Singing Voice Synthesis
Yizhong Geng, Tian-Hao Zhang, Chunfeng Wang +7
Reference-conditioned melody-preserving lyric editing replaces words while retaining a performance's timing, singer identity, and naturalness. Continuous-latent autoregression avoi…
Stable Curves, Unstable Items: Item-Level Scaling Heterogeneity in Video LLMs
Wenzhang Sun, Chunfeng Wang, Xiangchen Yin +3
Aggregate scaling curves suggest that Video LLMs improve smoothly or saturate as visual budgets grow. We show that this view can conceal large, opposing changes at the item level.…
MeloCodec: Harnessing Melodic Priors for High-Fidelity Singing Voice Representation
Yizhong Geng, Wenxin Fu, Kecan Mao +7
Neural audio codecs serve as fundamental tokenizers for LLM-based audio generation. While semantic priors are widely exploited to enhance linguistic intelligibility, the integratio…
Holo-World: Unified Camera, Object and Weather Control for Video World Model
Xiangchen Yin, Wenzhang Sun, Jiahui Yuan +6
Video world models are moving toward preserving an observed world under controllable camera and object motion while allowing its environmental state to change. Yet these controls r…
Preserve, Reveal, Expand: Towards Faithful 4D Video Editing with Region-Aware Conditioning and Benchmarking
Zhangchi Hu, Wenzhang Sun, Xiangchen Yin +5
While 4D-driven video diffusion models can generate visually plausible videos, faithful 4D editing poses a stricter requirement: source-observed content should be preserved, while…
MUSE: A Multi-agent Framework for Unconstrained Story Envisioning via Closed-Loop Cognitive Orchestration
Wenzhang Sun, Zhenyu Wang, Zhangchi Hu +3
Generating long-form audio-visual stories from a short user prompt remains challenging due to an intent-execution gap, where high-level narrative intent must be preserved across co…
Multi-Loss Learning for Speech Emotion Recognition with Energy-Adaptive Mixup and Frame-Level Attention
Cong Wang, Yizhong Geng, Yuhua Wen +7
Speech emotion recognition (SER) is an important technology in human-computer interaction. However, achieving high performance is challenging due to emotional complexity and scarce…
DrivingScene: A Multi-Task Online Feed-Forward 3D Gaussian Splatting Method for Dynamic Driving Scenes
Qirui Hou, Wenzhang Sun, Chang Zeng +3
Real-time, high-fidelity reconstruction of dynamic driving scenes is challenged by complex dynamics and sparse views, with prior methods struggling to balance quality and efficienc…
PAGS: Priority-Adaptive Gaussian Splatting for Dynamic Driving Scenes
Ying A, Wenzhang Sun, Chang Zeng +3
Reconstructing dynamic 3D urban scenes is crucial for autonomous driving, yet current methods face a stark trade-off between fidelity and computational cost. This inefficiency stem…
MotionStrata: Hierarchical Motion Latents for Compact Video Autoencoding
Wenzhang Sun, Huaize Liu, Chunfeng Wang +3
First-frame-conditioned video autoencoders represent a clip with persistent content and a compact motion code. Although this removes much of the appearance redundancy, the remainin…