WavLink: Compact Audio-Text Embeddings with a Global Whisper Token

• URL: http://arxiv.org/abs/2601.15118v2
• Date: Thu, 22 Jan 2026 08:55:20 GMT
• Title: WavLink: Compact Audio-Text Embeddings with a Global Whisper Token
• Authors: Gokul Karthik Kumar, Ludovick Lepauloux, Hakim Hacid,
• Abstract summary: We present WavLink, a compact audio-text embedding model that augments Whisper encoder with a learnable global token.<n>Our two-stage training recipe across three model sizes, combined with Matryoshka-style supervision, improves scalability, enabling 8x smaller embeddings with minimal performance drop.
• Score: 4.000493292896401
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Whisper has become the de-facto encoder for extracting general-purpose audio features in large audio-language models, where a 30-second clip is typically represented by 1500 frame features projected into an LLM. In contrast, audio-text embedding models like CLAP-based models have largely relied on alternative audio encoders (e.g., HTS-AT, PaSST), and have not leveraged Whisper effectively. We present WavLink, a compact audio-text embedding model that augments Whisper encoder with a learnable global token, trained jointly with a text encoder. Through a systematic study of design choices, including pretrained text encoders, loss functions, training modes, and data mixtures, we identify configurations that yield state-of-the-art retrieval performance. Our two-stage training recipe across three model sizes, combined with Matryoshka-style supervision, improves scalability, enabling 8x smaller embeddings with minimal performance drop. WavLink also demonstrates competitive performance on AIR-Bench with MCQs and zero-shot classification.

Related papers

• Scaling Open Discrete Audio Foundation Models with Interleaved Semantic, Acoustic, and Text Tokens [62.56027815951259]
  Current audio language models are predominantly text-first, either extending pre-trained text LLM backbones or relying on semantic-only audio tokens.<n>This paper presents a systematic empirical study of native audio foundation models that apply next-token prediction to audio at scale.
  arXiv  Detail & Related papers  (2026-02-18T18:32:46Z)
• MAGIC-Enhanced Keyword Prompting for Zero-Shot Audio Captioning with CLIP Models [2.3310964423816896]
  Automated Audio Captioning (AAC) generates captions for audio clips but faces challenges due to limited datasets.<n>We propose a zero-shot AAC system that leverages pre-trained models, eliminating the need for extensive training.<n> Experimental results demonstrate a 35% improvement in NLG mean score (from 4.7 to 7.3) using MAGIC search with the WavCaps model.
  arXiv  Detail & Related papers  (2025-09-16T02:36:00Z)
• UniVerse-1: Unified Audio-Video Generation via Stitching of Experts [59.38012380516272]
  We introduce UniVerse-1, a unified, Veo-3-like model capable of simultaneously generating coordinated audio and video.<n>To enhance training efficiency, we bypass training from scratch and instead employ a stitching of experts (SoE) technique.
  arXiv  Detail & Related papers  (2025-09-07T17:55:03Z)
• Low Frame-rate Speech Codec: a Codec Designed for Fast High-quality Speech LLM Training and Inference [10.909997817643905]
  We present the Low Frame-rate Speech Codec (LFSC): a neural audio that leverages a finite scalar quantization and adversarial training with large speech language models to achieve high-quality audio compression with a 1.89 kbps and 21.5 frames per second. We demonstrate that our novel LLM can make the inference of text-to-speech models around three times faster while improving intelligibility and producing quality comparable to previous models.
  arXiv  Detail & Related papers  (2024-09-18T16:39:10Z)
• WavTokenizer: an Efficient Acoustic Discrete Codec Tokenizer for Audio Language Modeling [63.8735398698683]
  A crucial component of language models is the tokenizer, which compresses high-dimensional natural signals into lower-dimensional discrete tokens.<n>We introduce WavTokenizer, which offers several advantages over previous SOTA acoustic models in the audio domain.<n>WavTokenizer achieves state-of-the-art reconstruction quality with outstanding UTMOS scores and inherently contains richer semantic information.
  arXiv  Detail & Related papers  (2024-08-29T13:43:36Z)
• Enhancing Automated Audio Captioning via Large Language Models with Optimized Audio Encoding [30.46616330202622]
  Automated audio captioning (AAC) is an audio-to-text task to describe audio contents in natural language. Recent advancements in large language models (LLMs) have opened up possibilities for improving AAC. Our method obtains a 33.0 SPIDEr-FL score, outperforming the winner of DCASE 2023 Task 6A.
  arXiv  Detail & Related papers  (2024-06-19T07:09:46Z)
• Audio-Visual Generalized Zero-Shot Learning using Pre-Trained Large Multi-Modal Models [53.48409081555687]
  In this work, we explore such large pre-trained models to obtain features, i.e. CLIP for visual features, and CLAP for audio features. We propose a simple yet effective model that only relies on feed-forward neural networks. Our framework achieves state-of-the-art performance on VGGSound-GZSL, UCF-GZSL, and ActivityNet-GZSL.
  arXiv  Detail & Related papers  (2024-04-09T13:39:37Z)
• Text-Driven Foley Sound Generation With Latent Diffusion Model [33.4636070590045]
  Foley sound generation aims to synthesise the background sound for multimedia content. We propose a diffusion model based system for Foley sound generation with text conditions.
  arXiv  Detail & Related papers  (2023-06-17T14:16:24Z)
• CLIPSonic: Text-to-Audio Synthesis with Unlabeled Videos and Pretrained Language-Vision Models [50.42886595228255]
  We propose to learn the desired text-audio correspondence by leveraging the visual modality as a bridge. We train a conditional diffusion model to generate the audio track of a video, given a video frame encoded by a pretrained contrastive language-image pretraining model.
  arXiv  Detail & Related papers  (2023-06-16T05:42:01Z)
• Pre-Training Transformer Decoder for End-to-End ASR Model with Unpaired Speech Data [145.95460945321253]
  We introduce two pre-training tasks for the encoder-decoder network using acoustic units, i.e., pseudo codes. The proposed Speech2C can relatively reduce the word error rate (WER) by 19.2% over the method without decoder pre-training.
  arXiv  Detail & Related papers  (2022-03-31T15:33:56Z)
• Automatic Audio Captioning using Attention weighted Event based Embeddings [25.258177951665594]
  We propose an encoder-decoder architecture with light-weight (i.e. with lesser learnable parameters) Bi-LSTM recurrent layers for AAC. Our results show that an efficient AED based embedding extractor combined with temporal attention and augmentation techniques is able to surpass existing literature.
  arXiv  Detail & Related papers  (2022-01-28T05:54:19Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.