This week the 2020 International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2020) is taking place virtually. ICASSP is the world’s largest and most comprehensive technical conference focused on signal processing and its applications. It is also one of the premiere conferences for audio research worldwide. In recent years, machine learning approaches have made a significant impact at this conference, a result of the evolution of audio research and signal processing more broadly.
This year Adobe Research has made a strong showing at ICASSP with 11 accepted papers, including 8 from the recently formed Audio Research Group. The papers span a broad range of audio domains including speech, music, and environmental audio. They tackle a variety of problems including music similarity, sound event detection, voice conversion, source separation, and dialogue systems (to name some), and employ a wide range of machine learning and signal processing techniques.
Many of the accepted papers are the outcome of audio research internships. Visit our internships program page to learn more. For those interested, please watch for future internship application announcements.
We welcome you to connect with us about our papers, our audio research, or about research at Adobe in general. We can be reached via email or Twitter.
Adobe Research papers at ICASSP 2020:
- Acoustic Matching by Embedding Impulse Responses, Jiaqi Su, Zeyu Jin, and Adam Finkelstein.
- A Simple But Effective BERT Model for Dialog State Tracking on Resource-Limited Systems, Tuan Manh Lai, Quan Hung Tran, Trung Bui and Daisuke Kihara.
- Chirping Up The Right Tree: Incorporating Biological Taxonomies Into Deep Bioacoustic Classifiers, Jason Cramer, Vincent Lostanlen, Andrew Farnsworth, Justin Salamon and Juan Pablo Bello.
- Disentangled Multidimensional Metric Learning for Music Similarity, Jongpil Lee, Nicholas J. Bryan, Justin Salamon, Zeyu Jin and Juhan Nam.
- End-To-End Non-Negative Autoencoders For Sound Source Separation, Shrikant Venkataramani , Efthymios Tzinis and Paris Smaragdis.
- F0-Consistent Many-To-May Non-Parallel Voice Conversion Via Conditional Autoencoder, Kaizhi Qian, Zeyu Jin, Mark Hasegawa-Johnson and Gautham J. Mysore.
- Few-Shot Sound Event Detection, Yu Wang, Justin Salamon, Nicholas J. Bryan and Juan Pablo Bello.
- Impulse Response Data Augmentation and Deep Neural Networks for Blind Room Acoustic Parameter Estimation, Nicholas J. Bryan
- One-Shot Parametric Audio Production Style Transfer with Application to Frequency Equalization, Stylianos I. Mimilakis, Nicholas J. Bryan and Paris Smaragdis.
- Sound Event Detection in Synthetic Domestic Environments, Romain Serizel, Nicolas Turpault, Ankit Shah and Justin Salamon
- Two-Step Sound Source Separation: Training on Learned Latent Targets, Efthymios Tzinis, Shrikant Venkataramani, Zhepei Wang, Cem Subakan and Paris Smaragdis.
A list of all audio-related Adobe Research publications can be found here.