arXiv Computation and Language

Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMs

arXiv Computation and Language
Sep 25

Temporal Taxation Compounds Under Post-Training Compression of Whisper Models

The paper investigates how post‑training compression techniques—such as pruning, quantization, and distillation—affect demographic fairness in Whisper speech‑recognition models. It finds that pruning and INT4 quantization significantly widen word‑error‑rate gaps between demographic groups, especially for Black/AA and Asian speakers, while distillation tends to reduce these gaps. The study introduces a temporal‑taxation metric to quantify the increased correction effort required for marginalized speakers after compression.

By Srishti Ginjala, Eric Fosler-Lussier, Christopher W. Myers, Srinivasan Parthasarathy
arXiv Machine Learning
Sep 24

Six Layers Less: Encoder Pruning for Whisper with Label-Free Recovery

The paper introduces a method to prune six layers from the encoder of OpenAI’s Whisper ASR model, reducing the encoder stack by 18.5% without requiring custom inference code. Layers are selected based on their minimal impact on Word Error Rate when removed. After pruning, the model’s WER rises from 18.2% to 21.9%, but distillation with unlabeled monolingual speech data lowers it to 20.1%. "whyItMatters":"The approach offers a straightforward way to accelerate Whisper inference by simplifying the encoder while maintaining acceptable accuracy, and the released code and model enable immediate adoption by the community."

By Rasmus Aagaard, Nicki Skafte Detlefsen
arXiv AI
Sep 18

Scaling Audio Models Efficiently: Joint Optimization of Scale, Resolution, Adaptation, Precision, and Sparsity

The paper introduces a compression framework for the Whisper automatic speech recognition model that jointly optimizes six deployment dimensions—model size, temporal resolution, encoder token stride, low‑rank adaptation capacity, weight precision, and sparsity pattern—using NSGA‑III. The optimization targets three objectives: word error rate, inference FLOPs, and memory footprint. Evaluating 1,680 configurations, the study identifies compression combinations that outperform single‑axis scaling and notes that 1:4 structured sparsity cannot maintain acceptable accuracy within the tested budgets.

By Vyom Agarwal, Mokshda Gangrade, Siddharth Pal, Jerry Wu
arXiv Computation and Language
Oct 23

WaxalNet: Benchmarking and Adapting Speech Recognition for Spontaneous Speech in 19 African Languages

The paper introduces WaxalNet, a benchmark and adaptation study for spontaneous speech recognition across 19 African languages using the WAXAL corpus. It evaluates zero‑shot and fine‑tuned performance of several multilingual foundation models (Whisper, MMS, Omnilingual ASR, SeamlessM4T‑v2), showing that fine‑tuning can substantially reduce word error rates and that LLM decoders generally outperform CTC decoders when the encoder is shared. The authors also demonstrate that a single multilingual model can serve all languages with only a modest WER penalty and provide a native‑speaker audit of failure modes.

By Victor Tolulope Olufemi, Oreoluwa Babatunde, Ramsey Njema, Bolarinwa Gbotemi, Wanchi Lucia Yen, John Uzodinma, Sunday Ajayi, Oluwademilade Williams, Kausar Moshood, Innocent Elendu Anyaele, Akebert Arefaine, Candace Hunzwi, Wongel Dawit Daniel, Emmilly Namuganga, Cleophas Kadima, Athanase Bahizire, Onitsiky Ranaivoson, Emmanuel Aaron, Nicholaus Ladislaus, Idris Muhammed, Jonathan Enoch Simenya, Martin Koome, Matewos Tegete Endaylalu, Peter Ifeoluwa Adeyemo, Hondi Prisca Birindwa, Ukachi Agnes Eze-Mbey, Yacoba Oduro-Yeboah, Pericles Adjovi, Mikel K. Ngueajio, Toluwani Aremu, Prasenjit Mitra