AVENUE is a new benchmark and evaluation framework for audio‑video editing that includes 1,291 source clips and 7,957 editing instructions covering audio‑targeted, video‑targeted, and coupled edits. It introduces a sample‑specific, modality‑aware evaluation that specifies the intended change and the content that must remain unchanged. The study applies this framework to joint, sequential, and separate editing models, revealing that existing models often alter unintended modalities, highlighting a key challenge in controllable AV editing.
By Hayeon Kim, Yoojin Jang, Jaejun Yoo
arXiv:2609.08275v1 Announce Type: new
Abstract: Recent multi-shot audio-video generators can produce increasingly coherent and cinematic outputs, but coherence does not imply the ability to execute e...
By Tianyi Zeng, Junchao Liao, Yujie Wei, Ziying Zhang, Litao Li, Tianyi Wang, Zhichao Wei, Shuyao Xu, Wenwen Qiang, Siyu Zhu, Zhenghao Zhang, Long Qin
OmniEdit-Bench introduces a comprehensive benchmark for instruction-based video editing (IVE), addressing limitations of existing datasets by covering spatial, temporal, audio, and reference-based editing tasks and distinguishing explicit from implicit instructions. The evaluation framework assesses editing quality across accuracy, preservation, realism, and consistency, using human judgments and vision-language models, and incorporates an accuracy-aware penalty to ensure instruction fidelity. Experiments reveal that current IVE models perform poorly, highlighting the need for improved methods.
By Chenxuan Miao, Yutong Feng, Yi Lu, Yunfeng Yan, Donglian Qi, Shiwei Zhang, Yu Liu, Xi Chen, Hengshuang Zhao
Neural music editing requires a system to perform a requested change while preserving the musical content that should remain unchanged. Existing evaluations often lack a paired target edit, making the...
arXiv:2609.14344v1 Announce Type: cross
Abstract: Instruction-guided music editors typically process each request independently, limiting their ability to support workflows in which users progressive...
By Quoc-Huy Trinh, Minh-Van Nguyen, Debesh Jha
The paper introduces a Dual-Transformer architecture with Cross-Attention for multi-camera view recommendation, achieving a 56.60% Precision@0.5 on the TVMCE dataset, surpassing the previous best of 37.16%. The model separates temporal encoding of past frames from candidate view querying, and an ablation study shows the SwinV2 backbone yields 69.65% Precision@0.5. Fine‑tuning with as little as 20% of a target video improves precision, suggesting efficient personalization for specific editing styles.
By Josep Cabacas-Maso, Carles Ventura, Ismael Benito-Altamirano
The paper introduces MuseCPEval, the first comprehensive framework for assessing Music Context Preservation (MuseCP) in music editing systems. It defines four categories of music facets and provides fine‑grained metrics to detect subtle changes during editing tasks such as timbre transfer, instrument substitution, and genre transformation. The authors validate the metrics objectively and through a human study, and demonstrate their practical use in evaluating diverse editing systems, offering insights into each system’s strengths and limitations.
By Yash Vishe, Eric Xue, Xunyi Jiang, Zachary Novack, Junda Wu, Julian McAuley, Xin Xu
arXiv:2606. 08415v1 Announce Type: cross Abstract: While recent text-guided video editing models excel at elementary tasks (e.
By Jiangtao Wu, Jiaming Wang, Yiwen He, Yuanxing Zhang, Shihao Li, Dunyuan Liu, Xuedong Zhao, Jialu Chen, Zekun Moore Wang, Jiaheng Liu
Multi-camera systems are foundational to modern media production, and multi-camera editing is a critical task. This involves the proper selection of the appropriate camera view at each moment. In this...
arXiv:2606. 10627v1 Announce Type: cross Abstract: The quality of piano performance depends on nuanced timing, articulation, and dynamic control, but practice feedback is often summary-based and hard to act on.
By Kazuki Kawamura, Fujiki Nakamura, Hayato Nishioka, Momoko Shioki, Shinichi Furuya, Jun Rekimoto
SpanSynth-Edit is a flow‑matching model that enables MIDI‑guided synthesis and editing of multi‑instrument audio mixtures using low‑frame‑rate scalar‑quantised latents. It encodes instrument‑labelled note lifecycles as unordered event sets, pools them into a conditioning vector per audio‑latent frame, and uses contextual audio for instrument‑specific timbre guidance. The model supports editing by resynthesising target regions from revised MIDI and demonstrates competitive performance on single‑ and multi‑instrument benchmarks, including within‑frame onset control.
By Sungkyun Chang, Keshav Bhandari, Simon Dixon, Emmanouil Benetos
arXiv:2602. 03762v4 Announce Type: replace-cross Abstract: Visually-guided acoustic highlighting seeks to rebalance audio in alignment with the accompanying video, creating a coherent audio-visual experience.
By Hugo Malard, Gael Le Lan, Daniel Wong, David Lou Alon, Yi-Chiao Wu, Sanjeel Parekh