I'm very happy to see that MMAudio, which my talented colleagues (@mi141, A. Hayakawa, @yahshibu) and intern (@hkchengrex) at Sony AI have invested their time and effort into, is being tested by so many creative people in X arXiv:

1 spotlight, 5 posters from us at #ICLR2025 "Weighted Point Cloud Embedding for Multimodal Contrastive Learning Toward Optimal Similarity Metric" led by Toshimitsu Uesaka, with strong support from Prof. Taiji Suzuki (@btreetaiji), was selected as a spotlight. Congrats🎊

Fast sampling methods for discrete diffusion from our lab 🏎️ Our sampling schedule optimization method (Jump Your Steps) is accepted at #ICLR2025 Another (Di4C) is about distllation for discrete diffusion!

@eleonoragrassuc @iclr_conf @GiordanoCicchet @luigi_sigillo @dhan90001 @IspammL Congrats! Great work!

When you register for #ICASSP2025, don't forget to select our tutorial on diffusion models for audio: 🎶Transforming Chaos into Harmony: Diffusion Models in Audio Signal Processing🎶 See you in Hyderabad, India!🇮🇳

"これまでにevalaが発表してきた36の立体音響作品のサウンド・データを学習したサウンドエフェクト生成AIを用いて，空間的作品をつくる試みです．サウンド（evalaがシグネチャーとして作品の始まりに用いている汽笛の音）とテキスト（学習に用いられた作品のうち8作品のタイトルをチャンネルごとに選択）の二つをリファレンス（プロンプト）として，リアルタイムかつマルチチャンネルで「evalaのような音」が生成され続けます．このプロジェクトは，作家不在でも立体音響作品を永続的に継承・制作しうる新しいアーカイヴのかたちを探求する実験であり，本作品はその最初のスケッチとなります．"

RT @SonyAI_global: 🚀 PaGoDA by Sony AI: High-res image generation without retraining! Fast, efficient, and quality-focused. #NeurIPS2024 h…

RT @SonyAI_global: #GenWarp by Sony AI creates realistic perspectives from a single image! See how it works 🧵 #NeurIPS2024

🎶Large music models from our team🎶: 1. SoniDo🎼 for music mixing, demixing, transcription, etc. pdf: 2. OpenMU🧙‍♂️ for music captioning, reasoning, lyric understanding, etc. pdf: code: demo @ISMIRConf : #ISMIR2024

A blog on our two papers has been published

A list of diffusion works & tutorial (at #ISMIR2024) from our lab! [ML] #NeurIPS24 (GenWarp: Novel View Synthesis) #NeurIPS24 (PaGoDA: Multi-Scale 1 Step Generator) #ICLR24 (CTM: Fast Image Gen.) #ICLR24 (MPGD: Guided Diffusion) #ICML23 (FP-Diff: Consistency-type Model) #ICML23 (Blind Inverse) [Audio/NLP] #ACL24 (Knowledge Gen.) #IJCAI24 (Music Editing) #ICASSP24 (Declipping) #ICASSP24 (Speech Enh.) #ICASSP23 (Music Transcription) #ICASSP23 (Vocoder) #ICASSP23 (Dereverb) #INTERSPEECH23 (Speech Enh.)

Here's a sneak peek of a crowd-based competition on sounding video generation starting from Oct. 1st! #ECCV2024

If you are working on a family of inverse problems using diffusion models and are confused about their relationships, this survey paper will clear your head!

RT @SonyAI_global: Meet @mittu1204, Lead Research Scientist, overseeing music and sound research within our AI for Creators Flagship. Lear…

A list of diffusion works + tutorial from our lab! [ML] #ACL2024 (Knowledge Generation) #ICLR2024 (Consistency Trajectory Model) #ICLR2024 (Manifold Preserving Guided Diffusion) #ICML2023 (Consistency-type Model) #ICML2023 (Blind Image Restoration) [Audio] #IJCAI2024 (Music Editing) #ICASSP2024 (Declipping) #ICASSP2024 (Speech Enhancement) #ICASSP2023 (Music Transcription) #ICASSP2023 (Vocoder) #ICASSP2023 (Dereverb) #INTERSPEECH2023 (Speech Enhancement)

@zicokolter Congrats! 🎉

📢For researchers in the Audio-Visual field, the call for papers of our #ECCV2024 workshop (AVGenL) is out Don't miss the deadline July 15⏳

If you will be at #ismir2024, don't forget to register our tutorial led by ⁦⁦@JCJesseLai⁩