KeisukeImoto

"XACLE Challenge 2026: The First x-to-Audio Alignment Challenge" introduces a new challenge on aligning various input modalities with audio. ieeexplore.ieee.org/document/11464…

"SONAR: Self-Distilled Continual Pre-training for Domain Adaptive Audio Representation" looks at how audio foundation models can keep adapting to new domains while retaining what they have already learned. ieeexplore.ieee.org/document/11465…

I'll be at ICASSP 2026 next week! If you will also be there, I would be very happy to catch up and have a chat. We will also present the following 2 papers:

Volume 322 proceedings.mlr.press/v312 Proccedings of the AAAI Workshop on Audio-Centric AI Is now available on PMLR.

DCASE T-shirt history👕 Not all versions, but still many here!

Yizhou Zhang, Yuan Gao, Wangjin Zhou, Zicheng Yuan, Keisuke Imoto, Tatsuya Kawahara, "SONAR: Self-Distilled Continual Pre-training for Domain Adaptive Audio Representation," arxiv.org/abs/2509.15703

Our paper on continual self-supervised audio representation learning has been accepted to #ICASSP2026🎉 arxiv.org/abs/2509.15703

We have also released the ground truth scores on the test dataset for the first XACLE Challenge. We hope many of you will find them useful. github.com/XACLE-Challeng…

[For XACLE Challenge participants] We have released the results of the first XACLE Challenge. Thank you all for your participation. Invitations for paper submission to ICASSP 2026 will be sent to the top five ranked teams later. xacle.org/results.html

Thrilled to announce that our paper (Onoma-to-Wave) received the APSIPA Sadaoki Furui Prize Paper Award🎉 Huge thanks to my co-authors (especially Okamoto-sensei) and everyone who supported this project. apsipa.org/apsipa.htm

Our new continual pre-training method for general purpose audio SSL model! Hope you’ll enjoy it!

"Audio-Centric AI: Towards Real-World Multimodal Reasoning and Application Use Cases" has been accepted as an #AAAI2026 workshop🎉 We're looking forward to seeing your contributions, with submissions due by 24 October 2025. #AAAI @RealAAAI

"Audio-Centric AI: Towards Real-World Multimodal Reasoning and Application Use Cases" has been accepted as an AAAI 2026 workshop🎉 We're looking forward to seeing your contributions, with submissions due by 24 October 2025. sites.google.com/view/audio-aaa…

Three of our papers on anomalous sound detection have been accepted to the DCASE 2025 Workshop🎉 Many thanks to all my co-authors!

The First XACLE Challenge has been accepted as the ICASSP2026 SP Grand Challenge🎉 @ieeeICASSP

Our paper "Context-Aware Query Refinement for Target Sound Extraction: Handling Partially Matched Queries" has been accepted to IEEE WASPAA 2025!🎉 Huge thanks to @KeisukeImoto for incredible support. #WASPAA2025

Our paper got accepted to IEEE WASPAA 2025! Huge thanks and congrats to my co-authors🎉

📢 The results for most of the tasks of DCASE 2025 challenge are out! dcase.community/challenge2025/

We are holding a competition for the automatic evaluation of text-to-audio generation. We will evaluate the semantic alignment between text and audio. Please join us! Pre-registration:forms.gle/s5pNf72Z7KWCfS… *Pre-registration does not obligate you to participate in competition

We are happy that🦉 OWLS, 18B to 0.25B open ASR/AST limited data scaling laws, has been accepted to @icmlconf 2025 led by @chenwanch1 @MXzBFhjFpS1jyMI from @shinjiw_at_cmu @WavLab and @NVIDIAAI Models: huggingface.co/collections/es… Paper: arxiv.org/pdf/2502.10373 Deepspeed ESPNet: github.com/espnet/espnet