Publications

Unifying Global and Near-Context Biasing in a Single Trie Pass

I. Thorbecke, E. Villatoro-Tello, J. P. Zuluaga, S. Kumar, S. Burdisso, P. Rangappa, A. Carofilis, S. Madikeri, P. Motlicek, K. Pandia, K. Hacioglu, A. Stolcke.

Single-pass trie unifies global vocabulary biasing with utterance-level context biasing for transducer ASR.

Speech Data Selection for Efficient ASR Fine-Tuning

P. Rangappa, S. Madikeri, J. P. Zuluaga, J. Villatoro-Tello, P. Motlicek.

A domain classifier plus pseudo-label filtering cuts ASR fine-tuning compute by ~40% at matched WER.

XLSR-Transducer: Streaming ASR for Self-Supervised Pretrained Models

S. Madikeri, J. P. Zuluaga, P. Rangappa, J. Villatoro-Tello, P. Motlicek.

Streaming ASR atop a frozen self-supervised backbone, without sacrificing non-streaming accuracy.

TokenVerse: Towards Unifying Speech and NLP Tasks via Transducer-based ASR

S. Kumar, S. Madikeri, J. P. Zuluaga, I. Thorbecke, E. Villatoro-Tello, S. Burdisso, P. Motlicek, K. S, A. Ganapathiraju.

A single transducer framework that jointly handles ASR, speaker change, and named-entity tagging — one model, one decoding pass.

Fast Streaming Transducer ASR Prototyping via Knowledge Distillation with Whisper

I. Thorbecke, J. P. Zuluaga, E. Villatoro-Tello, S. Kumar, P. Rangappa, S. Burdisso, P. Motlicek, K. S, A. Ganapathiraju.

Distill Whisper into a streaming transducer in hours instead of days — practical recipe for low-latency, low-resource ASR.

LM-assisted keyword biasing with Aho-Corasick algorithm for Transducer-based ASR

I. Nigmatulina, S. Madikeri, E. Villatoro-Tello, P. Motlicek, J. P. Zuluaga, K. Pandia, A. Ganapathiraju.

Aho-Corasick keyword spotting fused with an LM during transducer decoding — boosts rare-word recognition without retraining.

Contextual Biasing for Keyword Recognition on Low-Resource Air Traffic Control Data

M. Kocour, K. Veselý, I. Szoke, S. Kesiraju, J. P. Zuluaga, A. Blatt, M. Motlíček, J. Černocký.

Contextual biasing with deep fusion for domain-specific keyword recognition in ATC.

Open-source Conversational AI with SpeechBrain 1.0

M. Ravanelli, T. Parcollet, A. Moumen, S. de Langen, C. Subakan, P. Plantinga, Y. Liao, S. Cornell, D. Roman, S. Moradi, D. Chander, D. Petermann, Y. Wang, J. P. Zuluaga, et al.

Co-authored the 1.0 release of SpeechBrain — a PyTorch toolkit for conversational AI.

End-to-end single-channel speaker-turn aware conversational speech translation

J. P. Zuluaga, Z. Huang, X. Niu, R. Paturi, S. Srinivasan, P. Mathur, B. Thompson, M. Federico.

First end-to-end speech translation system that handles speaker turns and overlapped speech on a single channel.

HyperConformer: Multi-Head HyperMixer for Efficient Speech Recognition

F. Mai, J. P. Zuluaga, T. Parcollet, P. Motlicek.

Replaces Conformer attention with HyperMixer, matching accuracy at a fraction of the compute.

Implementing Contextual Biasing in GPU Decoder for Online ASR

I. Nigmatulina, S. Madikeri, E. Villatoro-Tello, P. Motlicek, J. P. Zuluaga, K. Pandia, A. Ganapathiraju.

Adds contextual biasing to a GPU CTC/attention decoder for online ASR with minimal latency overhead.

Effectiveness of Text, Acoustic, and Lattice-Based Representations in Spoken Language Understanding Tasks

E. Villatoro-Tello, S. Madikeri, J. P. Zuluaga, B. Sharma, S. Sarfjoo, I. Nigmatulina, P. Motlicek, A. Ivanov, A. Ganapathiraju.

Systematic comparison of text, acoustic, and lattice features for SLU — lattices recover information lost in 1-best ASR.

A Virtual Simulation-Pilot Agent for Training of Air Traffic Controllers

J. P. Zuluaga, A. Prasad, I. Nigmatulina, S. Sarfjoo, P. Motlicek, M. Kleinert, H. Helmke, O. Ohneiser, Q. Zhan.

End-to-end virtual pilot agent — ASR, intent parsing, and TTS — for training and evaluating air traffic controllers.

Lessons Learned in ATCO2: 5000 hours of Air Traffic Control Communications for Robust Automatic Speech Recognition and Understanding

J. P. Zuluaga, K. Vesely, A. Blatt, P. Motlicek, et al.

Postmortem of the ATCO2 project — what worked, what didn't, and what's next for low-resource ASR in safety-critical domains.

CommonAccent: Exploring Large Acoustic Pretrained Models for Accent Classification

J. P. Zuluaga, S. Sarfjoo, A. Prasad, I. Nigmatulina, P. Motlicek, K. Ondrej, O. Ohneiser, H. Helmke.

Accent classification benchmark on Common Voice using large self-supervised models — **Best Student Paper nominee**.

ATCO2 Corpus: A Large-Scale Dataset for Research on Automatic Speech Recognition and Natural Language Understanding of Air Traffic Control Communications

J. P. Zuluaga, K. Vesely, I. Szoke, P. Motlicek, M. Kocour, M. Rigault, K. Choukri, A. Prasad, S. Sarfjoo, I. Nigmatulina, C. Cevenini, P. Kolcarek, A. Tart, J. Cernocky, D. Klakow.

5,000 hours of Air Traffic Control communications — the largest open ATC speech corpus.

How Does Pre-Trained Wav2Vec 2.0 Perform on Domain Shifted ASR? An Extensive Benchmark on Air Traffic Control Communications

J. P. Zuluaga, A. Prasad, I. Nigmatulina, S. Sarfjoo, P. Motlicek, M. Kleinert, H. Helmke, O. Ohneiser, Q. Zhan.

Systematic study of self-supervised pretraining under domain shift — 20–40% relative WER cut on Air Traffic Control.

BERTraffic: BERT-based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications

J. P. Zuluaga, S. Sarfjoo, A. Prasad, I. Nigmatulina, P. Motlicek, K. Vesely, M. Kocour, I. Szoke.

Joint speaker-role and speaker-change detection for ATC using BERT — 27% DER reduction over audio-only baselines.

A Two-Step Approach to Leverage Contextual Data: Speech Recognition in Air Traffic Control

J. P. Zuluaga, I. Nigmatulina, A. Prasad, P. Motlicek, K. Vesely, M. Kocour, I. Szoke.

IDIAPers @ Causal News Corpus 2022: Efficient Causal Relation Identification Through a Prompt-Based Few-Shot Approach

S. Burdisso, J. P. Zuluaga, E. Villatoro-Tello, M. Fajcik, M. Singh, P. Smrz, P. Motlicek.

Few-shot prompt-based causal relation extraction — winner-tier system on the Causal News Corpus 2022 shared task.

IDIAPers @ Causal News Corpus 2022: Extracting Cause-Effect-Signal Triplets via Pre-trained Autoregressive Language Model

M. Fajcik, M. Singh, J. P. Zuluaga, E. Villatoro-Tello, S. Burdisso, P. Motlicek, P. Smrz.

Generative LM for joint cause-effect-signal triple extraction — companion submission to the same Causal News shared task.

Speech and Natural Language Processing Technologies for Pseudo-Pilot Simulator

J. P. Zuluaga, A. Prasad, I. Nigmatulina, P. Motlicek, M. Kleinert, H. Helmke, O. Ohneiser, Q. Zhan.

Speech + NLP pipeline that drives a pseudo-pilot simulator for ATCO training — built on top of the ATCO2 corpus.

Contextual Semi-Supervised Learning: An Approach to Leverage Air-Surveillance and Untranscribed ATC Data in ASR Systems

J. P. Zuluaga, I. Nigmatulina, A. Prasad, P. Motlicek, K. Vesely, M. Kocour, I. Szoke.

Boosting of Contextual Information in ASR for Air-Traffic Call-Sign Recognition

I. Nigmatulina, R. Braun, J. P. Zuluaga, P. Motlicek.

Grammar Based Identification Of Speaker Role For Improving ATCO And Pilot ASR

A. Prasad, J. P. Zuluaga, P. Motlicek, S. Sarfjoo, I. Nigmatulina, K. Vesely, M. Kocour, I. Szoke.

Improving Callsign Recognition with Air-Surveillance Data in Air-Traffic Communication

I. Nigmatulina, R. Braun, J. P. Zuluaga, P. Motlicek.

Domain-Adversarial Based Model with Phonological Knowledge for Cross-Lingual Speech Recognition

Q. Xu, Y. Li, J. Shen, J. Liu, Y. Yang, J. P. Zuluaga.

Automatic Call Sign Detection: Matching Air Surveillance Data with Air Traffic Spoken Communications

J. P. Zuluaga, M. Kocour, K. Vesely, I. Szoke, P. Motlicek, M. Rigault, K. Choukri, A. Tart, J. Cernocky.

Automatic Speech Recognition Benchmark for Air-Traffic Communications

J. P. Zuluaga, P. Motlicek, Q. Zhan, K. Vesely, R. Braun.

Pkwrap: a PyTorch Package for LF-MMI Training of Acoustic Models

S. Madikeri, S. Tong, J. P. Zuluaga, A. Vyas, P. Motlicek, H. Bourlard.

PyTorch package for LF-MMI acoustic model training, bridging Kaldi and modern PyTorch workflows.

A CNN-based Methodology for Breast Cancer Diagnosis using Thermal Images

J. P. Zuluaga, Z. Al Masry, K. Benaggoune, S. Meraghni, N. Zerhouni.

A Survey of Breast Cancer Screening Techniques: Thermography and Electrical Impedance Tomography

J. P. Zuluaga, Z. Al Masry, K. Benaggoune, S. Meraghni, N. Zerhouni.

A Portable Breast Cancer Detection System Based on Smartphone with Infrared Camera

J. P. Zuluaga, Z. Al Masry, N. Zerhouni.