Synthetic Data for Data Efficient Speech, Audio and Natural Language Processing

Overview

Below are our most recent research on generating effective sythetic data for labelled-data effecient speech, audio and natural language processing.

Project Conference/Journal Year
LipGER: Visually-Conditioned Generative Error Correction for Robust Automatic Speech Recognition InterSpeech 2024
MeLFusion: Synthesizing Music from Image and Language Cues using Diffusion Models CVPR 2024
AV-RIR: Audio-Visual Room Impulse Response Estimation CVPR 2024
Listen2Scene: Interactive material-aware binaural sound propagation for reconstructed 3D scenes IEEE VR 2024
CompA: Addressing the Gap in Compositional Reasoning in Audio-Language Models ICLR 2024
AdVerb: Visually Guided Audio Dereverberation ICCV 2023
MESH2IR: Neural Acoustic Impulse Response Generator for Complex 3D Scenes ACM Multimedia 2022
FAST-RIR: Fast neural diffuse room impulse response generator ICASSP 2022
NeuralSound: Learning-based Modal Sound Synthesis With Acoustic Transfer SIGGRAPH 2022
GWA: A Large Geometric-Wave Acoustic Dataset for Audio Processing SIGGRAPH 2022
Point-based Acoustic Scattering for Interactive Sound Propagation via Surface Encoding IJCAI 2021
Learning Acoustic Scattering Fields for Dynamic Interactive Sound Propagation IEEE VR 2021
Scene-Aware Audio Rendering via Deep Acoustic Analysis TVCG 2020