Ishwarya Ananthabhotla
Machine Learning and Audio Research Scientist
About Me
I am a Member of Technical Staff, Audio at LiquidAI. I am interested in machine learning applied to topics in audio, with a special interest in spatial and multichannel audio, room acoustics, and egocentric understanding. I was formerly a Research Scientist at Meta Reality Labs Research Audio for 4.5 years.
I completed my Bachelors from MIT in 2015, my Masters from MIT in 2016, and my PhD from the MIT Media Lab in 2021, advised by Prof. Joe Paradiso. I was supported by the NSF Graduate Research Fellowship (2016-2019) and the Apple AI/ ML Fellowship (2020-2022). Outside of research, I enjoy singing, playing instruments, and writing; while at MIT, I also enjoyed DJ-ing at the WMBR radio station.
Publications
2026SPHERE: Automatic Music Upmixing via Audio Language Model Post-Training with Spatial Heuristic RewardsEMNLP[paper]
2026Ambisonics Encoding of Room Impulse Responses using a Device-Agnostic Diffusion ModelIWAENC[paper]
2026Learning Ego-Exo Visual Representations for Conversational Gaze Estimation (Best Presentation)CVPR Workshop[paper]
2026The RL-R Chat Dataset: Egocentric Conversations among Familiar Interlocutors for Multi-Modal Hearing Augmentation TechnologyICASSP[paper]
2025EgoAdapt: Adaptive Multisensory Distillation and Policy Learning for Efficient Egocentric PerceptionICCV[paper]
2024On HRTF Notch Frequency Prediction Using Anthropometric Features and Neural NetworksICASSP[paper]
2023A Two‑Dimensional Threshold Test for Reverberation Time and Direct‑to‑Reverberant RatioI3DA[paper]
2023Autonomous Room Acoustic Measurements Using Rapidly‑Exploring Random Trees and Gaussian ProcessesForum Acusticum[paper]
2022Towards the Prediction of Perceived Room Acoustical Similarity (Best Student Paper)AES AVAR[paper]
2021Cognitive Audio Interfaces: Mediating Sonic Information with an Understanding of How We HearIEEE Pervasive Computing[paper]
2021A Framework for Designing Head‑related Transfer Function (HRTF) Distance Metrics that Capture Localization PerceptionJASA Express Letters[paper]
2020Using a Neural Network Codec Approximation Loss to Improve Source Separation Performance in Limited Capacity NetworksIJCNN[paper]
2019Towards a Perceptual Loss: Using a Neural Network Codec Approximation as a Loss for Generative Audio ModelsACM Multimedia[paper]
2019HCU400: An Annotated Dataset for Exploring Aural Phenomenology through Causal UncertaintyICASSP[paper]
2018SoundSignaling: Realtime, Stylistic Modification of a Personal Music Corpus for Information DeliveryACM IMWUT[paper]
2017VisualSoundtrack: An Approach to Style Transfer in the Context of Soundtrack CompositionICMC[paper]
Music, Writing, Radio
I enjoy learning and singing Carnatic music and lighter forms of Indian music, as well as playing keys, harmonium, naal, and dholak. I also enjoy dabbling in song writing, music production, and mixing and mastering in my spare time. Some performances and recordings across genres: Sruthi Laya TCD 2014, MIT C-Show 2016, Someone Like You (Arr.) 2015. I also love writing short stories and poetry; some of my writing was awarded an MIT Karmel Prize, and I've written a few plays (An Untimely Life, A Familiar Melody) that have aired on my WMBR production, "The Daydream Company." You can sample some content from my days as a radio host below.
Post-It Wall: "If you could leave one piece of advice behind for the world to hear, what would it be?" Episode 3 - Guest: Neil Gaikwad, April 5th, 2017 Episode 4 - Guest: JJ Hernandez, April 19th, 2017 Episode 7 - Guest: Ravi Yegya-Raman, June 14th, 2017 Episode 9 - Guest: Arthi Vezhavendan, July 26th, 2017 Episode 10 - Guest: Vimala Nandula, August 9th, 2017 Episode 12 - Guest: Elizabeth Devine, September 20th, 2017 Of Moments and Melodies: "A whirlwind tour of the tunes and tales of the Indian sub-continent." Episode 2 - March 21nd, 2018 Episode 3 - April 4th, 2018 Episode 4 - April 18th, 2018 Episode 5 - May 2nd, 2018 Episode 6 - May 16th, 2018