@NicholasJBryan
Account based inUnited States
About this account
- Account based in
- United States
- Connected via
- United States App Store
Account-level information from X, not a live location or the device used for a specific post.
Head of Music AI / Principal Scientist, Adobe Research (personal account)
Joined April 2010
- Tweets101
- Following529
- Followers1.4K
- Likes1.1K
Pinned Tweet
Congrats @yblin98 and #team on the @NeurIPSConf 2026 Accept!
SOTA synchronized video-to-music generation without using any paired video-music training data! 👏👏👏
🎵🎵What if we could generate video soundtracks without paired video–music data? Introducing V2M-Zero, a method that generates music synchronized with video events. arxiv.org/abs/2603.11042 w. @CasebeerJonah @mtlong_88 @aniruddha26398 @gberta227 @NicholasJBryan
Excited for "Beats" is released in @Adobe #PremiereMobile on iOS!
We detect musically meaningful markers within your music to sync, snap, and cut to your video. Three visualization options corresponding to less → more markers.
w/@JojoGiltsoff, Eric Brandt, @j_p_caceres ++!
Introducing #AIStudio via #AdobeStock — edit stock assets before licensing stock.adobe.com/ai-studio
AI Studio features #AudioMatch -- create custom soundtracks for video before licensing.
Proud of the #team 🥂👇 + Adobe Stock's new economic models to blend #GenAI w/licensing.
🎶 V2M-Zero: SOTA time-sync'd video-to-music! 🎶
* Music sync'd to dance, scene cuts,
* Easily adapt existing TTM models,
* No paired video-music data,
* SOTA objective + human preference.
Exceptional work @yblin98 + @casebeer! w/@mtlong_88 @aniruddha26398 @gberta227, me
🎵🎵What if we could generate video soundtracks without paired video–music data? Introducing V2M-Zero, a method that generates music synchronized with video events. arxiv.org/abs/2603.11042 w. @CasebeerJonah @mtlong_88 @aniruddha26398 @gberta227 @NicholasJBryan
Audio VAEs + VQ-VAEs designed for #GenAI!
* Ultra-fast encoding for on-the-fly training pipelines,
* ~2x more compression (13Hz) w/frontier quality,
* Any format (mono, stereo LR, MS, mel, raw),
* Cont. or discrete latents.
👏 @CasebeerJonah! w/@__gzhu__ @zhepeiw03, me
GenAE: An audio autoencoder engineered for generative modeling.
To appear at ICASSP 2026.
w/ @__gzhu__ @zhepeiw03 @NicholasJBryan
arXiv: arxiv.org/abs/2602.15749
Video: youtu.be/gDIIuLb0cf0
TAC: Timestamped Audio Captioning 👇
This is big. SOTA audio reasoning. SOTA video reasoning. SOTA audio captioning. SOTA sound event detection. Better than Gemini. Better than Qwen.
TAC: Timestamped Audio Captioning
📑 paper: lnkd.in/getEz5xU
🌐 website with more demos: lnkd.in/gdw5TTuS
Thrilled about Stemphonic!
All-at-once Flexible Multi-stem Music Generation!
w/@slseanwu @__gzhu__ @j_p_caceres @huangcza and myself
Excited to announce our ICASSP 2026 paper "Stemphonic: All-at-once Flexible Multi-stem Music Generation" !
w/ @__gzhu__, @j_p_caceres, @huangcza, and @NicholasJBryan
🔊Demo stemphonic-demo.vercel.app
📰Paper arxiv.org/abs/2602.09891
More details in🧵
Nicholas J. Bryan retweeted
I've been playing with Generate Soundtrack a lot today. It is pretty great, and really fast.
I've been collecting a bunch of alternate music options from my videos.
Nicholas J. Bryan retweeted
Why no one is talking about the new Meta AI - Image and Video generation which is partnered with Midjourney and Black Forest Labs??
The generations are fascinating.
Image and Video generated with @Meta @AIatMeta
Music - @Adobe Generate Soundtrack
@alexandr_wang #MetaAI
Nicholas J. Bryan retweeted
Added sound in @AdobeFirefly , generate soundtrack feature
🎵Hiring summer interns on AI music 2026! @AdobeResearch adobe.ly/4dS1zc6
Past intern papers:
arxiv.org/abs/2504.15217 TMLR arxiv.org/abs/2507.07867 MLSP arxiv.org/abs/2410.05167 ICLR arxiv.org/abs/2403.10493 SPL arxiv.org/abs/2401.12179 ICML
arxiv.org/abs/2311.07069 TASLP
Nicholas J. Bryan retweeted
Thrilled to announce “MIDI-LLM: Adapting LLMs for Text-to-MIDI Music Generation” w/ @huangcza and Yoon Kim!
🎸 Live Demo midi-llm-demo.vercel.app
💻 github.com/slSeanWU/MIDI-LLM
🤗 huggingface.co/slseanwu/MIDI…
From a text prompt, it generates MIDIs you can edit directly in a DAW 🧵
Congrats @JCJesseLai and Team!
Tired to go back to the original papers again and again? Our monograph: a systematic and fundamental recipe you can rely on!
📘 We’re excited to release 《The Principles of Diffusion Models》— with @DrYangSong, @gimdong58085414, @mittu1204, and @StefanoErmon.
It traces the core ideas that shaped diffusion modeling and explains how today’s models work, why they work, and where they’re heading.
🧵You’ll find the link and a few highlights in the thread.
We’d love to hear your thoughts and join some discussions!
⚡ Stay tuned for our markdown version, where you can drop your comments!
Nicholas J. Bryan retweeted
Adobe’s new AI audio tools can add soundtracks and voice-overs to videos theverge.com/news/807809/ado…
Adobe's #GenerateSoundtrack is LIVE today! 🎉
Studio-quality music for storytellers🎵
* Trained on #Licensed data,
* Commercially safe, royalty-free, and cleared for any use, &
* Exported with #ContentCredentials for transparency and attribution.
Get started: firefly.adobe.com/generate/s…
Powered by the #FireflyAudioModel, and built by an incredible R&D team: @j_p_caceres @CasebeerJonah @__gzhu__ @zhepeiw03 @ailiefraser @NicholasJBryan
Excited for the Team. Much more to come!
#Adobe #AdobeMAX #GenerateSoundtrack #FireflyAudioModel
Introducing "DRAGON: Distributional Rewards Optimize Diffusion Generative Models"!
📖: arxiv.org/abs/2504.15217
🎹: ml-dragon.github.io/web/
A new framework for fine-tuning gen models towards a target distribution.
By Yatong Bai w/@CasebeerJonah @somayeh_sojoudi @NicholasJBryan
With an appropriate exemplar set, DRAGON achieves a 60.95% human-voted music quality win rate without training on human preference annotations.