The Second Workshop on Audio-Centric AI: Auditory Reasoning, Multimodal Understanding, Audio Generation, and Trustworthy Deployment (Audio-AAAI2)

About

This second Audio-AAAI workshop (Audio-AAAI2) explores recent and emerging directions in audio-centric AI, spanning auditory reasoning, multimodal scene understanding, audio generation, and trustworthy evaluation and deployment. It aims to connect researchers across audio, speech, and multimodal AI to identify shared technical challenges, strengthen evaluation practices, and foster collaboration on real-world applications.

Topics 

The topics of interest include, but are not limited to:

- Audio foundation models, audio-language models, and spoken language models.

- Auditory reasoning: audio question answering, audio retrieval-augmented generation, long-context audio understanding, and temporal grounding.

- Audio-visual-language understanding, audio-visual grounding, and auditory world models.

- Sound event detection and localization, acoustic scene analysis, and anomalous sound detection.

- Spatial audio, microphone array and virtual microphone techniques, and source separation.

- Robust audio and multimodal scene analysis under real-world constraints (e.g., asynchronous microphones, reverberant environments, audio-visual desynchronization).

- Audio and multimodal perception for robotics and embodied agents.

- Expressive and controllable generation of speech, music, and environmental sound.

- Video-conditioned generation (video-to-audio, dubbing, avatars) and audio-visual synchronization.

- Audio reasoning benchmarks, perceptual quality assessment, metrics, and efficient evaluation frameworks for multimodal reasoning and generation.

- Benchmarking dataset creation, sharing, and reproducibility across audio, speech, and video modalities.

- Trustworthy audio AI: deepfake detection, watermarking and provenance, privacy-preserving audio sensing, and efficient deployment on edge and wearable devices.

Submission Requirements

We invite full papers of up to 8 pages and short papers of up to 4 pages, excluding references. Submissions must use the AAAI-27 formatting template (https://aaai.org/conference/aaai/aaai-27/main-technical-track-call/) and be anonymized for double-blind review. Papers may be submitted under either the archival track or the non-archival track. Non-archival track papers will not be included in the proceedings and may consist of previously accepted work, preliminary studies or work under review. Archival track submissions must present original work that has not been published or submitted elsewhere. The Microsoft CMT service was used for managing the peer-reviewing process for this conference. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.

Format of Workshop

A one-day, in-person workshop combining invited talks, paper presentations, and interactive discussions to balance technical depth with participant engagement.

Attendance

Authors of accepted papers are invited to present. Participation is open to all interested individuals registered for the workshop through AAAI-27. No specific attendance cap is planned, subject to venue capacity. We expect 40-60 attendees.

Important Dates

•  20 November 2026: Paper Submission 

• 22 or 23 February 2027: Workshop Program

Workshop Committee 

• Nancy F. Chen, Center for Frontier AI Research, A*STAR, Singapore, Nancy_Chen@a-star.edu.sg

• Xiaoxue Gao, The Chinese University of Hong Kong, Shenzhen, China, gaoxiaoxue@cuhk.edu.cn

• Gordon Wichern, Mitsubishi Electric Research Laboratories, wichern@merl.com

• Keisuke Imoto, Kyoto University, Japan, keisuke.imoto@ieee.org

• Tatsuya Komatsu, LY Corporation, Japan, komatsu.tatsuya@lycorp.co.jp