Publications

Lifted Residual Score Estimation
This paper proposes two new techniques to im- prove the accuracy of score estimation. The first proposal is a new objective function …
Tejas Jayashankar, Jongha J. Ryu, Xiangxiang Xu, Gregory W. Wornell
Paper
Score-based Source Separation with Applications to Digital Communication Signals
We propose a new method for separating superimposed sources using diffusion-based generative models. Our method relies only on …
Tejas Jayashankar, Gary C.F. Lee, Alejandro Lancho, Amir Weiss, Yury Polyanksiy, Gregory Wornell
arXiv Link Openreview Link Project Webpage Code
Self-Supervised Representations for Singing Voice Conversion
A singing voice conversion model converts a song in the voice of an arbitrary source singer to the voice of a target singer. Recently, …
Tejas Jayashankar, Jilong Wu, Leda Sari, David Kant, Vimal Manohar, Qing He
Paper Audio Samples
System and Method for Detecting Adversarial Attacks
A linguistic system for transcribing an input, where the linguistic system comprises a processor configured to execute a neural network …
Jonathan Le Roux, Tejas Jayashankar, Pierre Moulin
Link
Architecture for Variable Bitrate Neural Speech Codec with Configurable Computation Complexity
Low bitrate speech codecs have become an area of intense research. Traditional speech codecs, which use signal processing methods to …
Tejas Jayashankar, Thilo Koehler, Kaustubh Kalgaonkar, Zhiping Xiu, Jilong Wu, Ju Lin, Prabhav Agrawal, Qing He
Paper
Image Compression using Sum-Product Networks
An estimated 79 zettabytes (1021 bytes) of data was generated worldwide in 2021 with even more data expected to be produced in the …
Tejas Jayashankar
Thesis
Detecting Audio Attacks on ASR Systems with Dropout Uncertainty
Various adversarial audio attacks have recently been developed to fool automatic speech recognition (ASR) systems. We here propose a …
Tejas Jayashankar, Jonathan Le Roux, Pierre Moulin
arXiv Link Conference Link
LAP-based Video Frame Interpolation
High-quality video frame interpolation often necessitates accurate motion estimation, which can be obtained using modern optical flow …
Tejas Jayashankar, Pierre Moulin, Thierry Blu, Chris Gilliam"
Paper