RT @mfu3ntes: 🎶📄Submissions are open for the 1st Latin American Music Information Retrieval Workshop (LAMIR). Submission deadline: Septembe….

I collaborated on a paper with some awesome ecologists that was just published in Trends in Ecology and Evolution. Check out @ellycknight's overview (and then check out the paper)!.

Also, NSI caption density has decreased by a factor of 3 in videos by high-valued studios in the past 10 yrs, possibly indicating increasing overreliance on ASR even within 'manual' captioning pipelines resulting in captions focused primarily on speech. Learn more in the paper!.

Our paper investigates the trends of NSI captioning in the past 10 years on YouTube. We find the presence of NSI captions is very low (~4% of videos) and when it is present, its density is also lower than expected.

Motivation: Most videos on YouTube are not required to be captioned according to any guidelines, and YouTube's automated captioning systems do not caption rich non-speech information (NSI) such as rich music, SFX, and paralinguistic info. So, how often is this info captioned?.

#CHI2024 On Wednesday at 11:45am in 321, I'm presenting "Unspoken Sound: Identifying Trends in Non-Speech Audio Captioning on YouTube" a paper in collaboration with @MayTheLloyd, @bick_jp, Khang Dang, @latent_audio, Jhanvi Pai, @mfu3ntes, and Sooyeon Lee.

#CHI2024 Excited that the SInC Lab has two papers at CHI this year. The first is today 2pm in 313a – Keita Ohshiro (@bick_jp) will be presenting his work "Audio Engineering by People Who Are deaf and Hard of Hearing: Balancing Confidence and Limitations"

We are recruiting survey participants to learn more about the accessibility of non-speech sound captions, in particular your experience in viewing non-speech captions in videos such as TV, YouTube, TikTok, Instagram, etc. Go to for the survey. Pls share!.

I have an opening in my lab at NJIT for a PhD student starting in Fall 2024. In particular, I'm looking for a student interested in human-AI collaboration for music composition, production, or performance. Get in touch if interested!.

RT @JonathanLeRoux: #SANE2023 is in just over 1 month, Thu 10/26 at NYU in Brooklyn!.Talk details are now up for @kchonyc, @YGongND, @huang….

RT @IEEE_WASPAA: We are working closely with the venue to maximize the number of additional guests. Plan to reopen the registration to the….

RT @lostanlen: I am hiring PhD students at @LaboLS2N 🦀 and @Ircam 🪗 on. 🦀 resource-efficient deep learning in the raw waveform . 🪗 birdsong….

RT @IEEE_WASPAA: WASPAA 2023 paper submission site is open! 👉Please submit your abstract by 4/21 and the final vers….

For those interested in environmental sound synthesis (e.g., for this year's #DCASE Foley challenge), check out a new paper led by Danzel Serrano (@dependanz). It presents a method for learning models for controllable sound synthesis, e.g. footsteps.

RT @IEEE_WASPAA: WASPAA 2023 calls for papers! The traditional intimate Mohonk Mountain House with exciting changes: double-blind review, a….

RT @justin_salamon: If you do music source separation or enhancement, consider participating in the 1st Cadenza challenge to improve the pe….

RT @huangcza: Passionate about creativity+music+art, HCI, and ML?. I’m still recruiting PhD and master students through @Mila_Quebec for Fa….

RT @youngmoo: .@functiontelechy @markbcartwright @InstrumentsLab @colinraffel @kmkinnaird @jcdevaney @asterix77 @drscotthawley . Hope you'l….

RT @arcticpenguin: Seeking postdoc who does quant or qual research in content moderation, online community management, hate speech online.….

RT @csteinmetz1: This is great for researchers and audio engineers! A platform for integrating your own deep learning models into plugins a….