WHISTRESS: Enriching Transcriptions with Sentence Stress Detection

TL;DR

WHISTRESS is a novel alignment-free method for sentence stress detection in transcriptions, trained on synthetic data, that outperforms existing approaches and generalizes well across benchmarks.

Contribution

Introduces WHISTRESS, an alignment-free sentence stress detection model trained on synthetic data, with strong zero-shot generalization capabilities.

Findings

Outperforms existing methods in stress detection accuracy.

Requires no additional priors during training or inference.

Demonstrates strong zero-shot generalization across benchmarks.

Abstract

Spoken language conveys meaning not only through words but also through intonation, emotion, and emphasis. Sentence stress, the emphasis placed on specific words within a sentence, is crucial for conveying speaker intent and has been extensively studied in linguistics. In this work, we introduce WHISTRESS, an alignment-free approach for enhancing transcription systems with sentence stress detection. To support this task, we propose TINYSTRESS-15K, a scalable, synthetic training data for the task of sentence stress detection which resulted from a fully automated dataset creation process. We train WHISTRESS on TINYSTRESS-15K and evaluate it against several competitive baselines. Our results show that WHISTRESS outperforms existing methods while requiring no additional input priors during training or inference. Notably, despite being trained on synthetic data, WHISTRESS demonstrates strong zero-shot generalization across diverse benchmarks. Project page: https://pages.cs.huji.ac.il/adiyoss-lab/whistress.