VOX-KRIKRI: Unifying Speech and Language through Continuous Fusion

• URL: http://arxiv.org/abs/2509.15667v1
• Date: Fri, 19 Sep 2025 06:42:42 GMT
• Title: VOX-KRIKRI: Unifying Speech and Language through Continuous Fusion
• Authors: Dimitrios Damianos, Leon Voukoutis, Georgios Paraskevopoulos, Vassilis Katsouros,
• Abstract summary: We present a framework that bridges pre-trained decoder-based large language models (LLM) and acoustic encoder-decoder architectures such as Whisper.<n>Our method operates fully in continuous text representation spaces, fusing Whisper's hidden decoder states with those of an LLM through cross-modal attention.
• Score: 7.96619533548369
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present a multimodal fusion framework that bridges pre-trained decoder-based large language models (LLM) and acoustic encoder-decoder architectures such as Whisper, with the aim of building speech-enabled LLMs. Instead of directly using audio embeddings, we explore an intermediate audio-conditioned text space as a more effective mechanism for alignment. Our method operates fully in continuous text representation spaces, fusing Whisper's hidden decoder states with those of an LLM through cross-modal attention, and supports both offline and streaming modes. We introduce \textit{VoxKrikri}, the first Greek speech LLM, and show through analysis that our approach effectively aligns representations across modalities. These results highlight continuous space fusion as a promising path for multilingual and low-resource speech LLMs, while achieving state-of-the-art results for Automatic Speech Recognition in Greek, providing an average $\sim20\%$ relative improvement across benchmarks.

Related papers

• LAMB: LLM-based Audio Captioning with Modality Gap Bridging via Cauchy-Schwarz Divergence [35.123477091633866]
  LAMB is an audio captioning framework that bridges the modality gap between audio embeddings and the text embedding space.<n>A Cross-Modal Aligner minimizes Cauchy-Schwarz divergence while maximizing mutual information.<n>A Two-Stream Adapter that extracts semantically enriched audio embeddings delivers richer information to the Cross-Modal Aligner.
  arXiv  Detail & Related papers  (2026-01-08T07:05:35Z)
• What Makes a Good Speech Tokenizer for LLM-Centric Speech Generation? A Systematic Study [58.55905182336196]
  Speech-language models (SLMs) offer a promising path toward unifying speech and text understanding and generation.<n>We investigate the role of speech tokenizer designs in LLM-centric SLMs, augmented by speech heads and speaker modeling.<n>We introduce multi-token prediction (MTP) into SLMs, enabling each hidden state to decode multiple speech tokens.
  arXiv  Detail & Related papers  (2025-06-14T15:26:31Z)
• DrVoice: Parallel Speech-Text Voice Conversation Model via Dual-Resolution Speech Representations [62.00227663434538]
  DRVOICE-7B establishes new state-of-the-art (SOTA) on OpenAudioBench and Big Bench Audio benchmarks.<n>This paper presents DrVoice, a parallel speech-text voice conversation model based on joint autoregressive modeling.
  arXiv  Detail & Related papers  (2025-06-11T02:57:22Z)
• TESU-LLM: Training Speech-LLMs Without Speech via Unified Encoder Alignment [15.899112804399193]
  We present textbfTESU-LLM, a novel framework that enables training speech-capable language models using only text data.<n>Our key insight is to leverage a unified encoder that maps semantically equivalent text and speech inputs to a shared latent space.<n>Despite being trained exclusively on text, TESU-LLM achieves strong performance on various speech-related benchmarks.
  arXiv  Detail & Related papers  (2025-06-01T09:27:55Z)
• Hearing and Seeing Through CLIP: A Framework for Self-Supervised Sound Source Localization [17.234696376137677]
  We introduce a framework that maps audios into tokens compatible with CLIP's text encoder, producing audio-driven embeddings.<n>These embeddings are used to generate sounding region masks, from which visual features are extracted and aligned with the audio embeddings.<n>Our findings show that alignment knowledge of pre-trained multimodal foundation model enables our method to generate more complete and compact localization for sounding objects.
  arXiv  Detail & Related papers  (2025-05-08T15:32:04Z)
• CAV-MAE Sync: Improving Contrastive Audio-Visual Mask Autoencoders via Fine-Grained Alignment [76.32508013503653]
  We propose CAV-MAE Sync as a simple yet effective extension of the original CAV-MAE framework for self-supervised audio-visual learning.<n>We tackle the mismatch between modalities by treating audio as a temporal sequence aligned with video frames, rather than using global representations.<n>We improve spatial localization by introducing learnable register tokens that reduce semantic load on patch tokens.
  arXiv  Detail & Related papers  (2025-05-02T12:59:58Z)
• Connecting Voices: LoReSpeech as a Low-Resource Speech Parallel Corpus [0.0]
  This paper introduces a methodology for constructing LoReSpeech, a low-resource speech-to-speech translation corpus.<n>LoReSpeech delivers both intra- and inter-language alignments, enabling advancements in multilingual ASR systems.
  arXiv  Detail & Related papers  (2025-02-25T14:00:15Z)
• Large Language Model Can Transcribe Speech in Multi-Talker Scenarios with Versatile Instructions [68.98811048970963]
  We present a pioneering effort to investigate the capability of large language models (LLMs) in transcribing speech in multi-talker environments.<n>We use WavLM and Whisper encoder to extract multi-faceted speech representations that are sensitive to speaker characteristics and semantic context.<n>Experiments reveal the promising performance of our proposed system, MT-LLM, in cocktail party scenarios.
  arXiv  Detail & Related papers  (2024-09-13T07:28:28Z)
• Improving Audio-Visual Speech Recognition by Lip-Subword Correlation Based Visual Pre-training and Cross-Modal Fusion Encoder [58.523884148942166]
  We propose two novel techniques to improve audio-visual speech recognition (AVSR) under a pre-training and fine-tuning training framework. First, we explore the correlation between lip shapes and syllable-level subword units in Mandarin to establish good frame-level syllable boundaries from lip shapes. Next, we propose an audio-guided cross-modal fusion encoder (CMFE) neural network to utilize main training parameters for multiple cross-modal attention layers.
  arXiv  Detail & Related papers  (2023-08-14T08:19:24Z)
• On decoder-only architecture for speech-to-text and large language model integration [59.49886892602309]
  Speech-LLaMA is a novel approach that effectively incorporates acoustic information into text-based large language models. We conduct experiments on multilingual speech-to-text translation tasks and demonstrate a significant improvement over strong baselines.
  arXiv  Detail & Related papers  (2023-07-08T06:47:58Z)
• Spoken Question Answering and Speech Continuation Using Spectrogram-Powered LLM [19.36630667212398]
  We present Spectron, a novel approach to adapting pre-trained large language models (LLMs) to perform spoken question answering (QA) and speech continuation. Key to our approach is a training objective that jointly supervises speech recognition, text continuation, and speech synthesis. Our method surpasses existing spoken language models in speaker preservation and semantic coherence.
  arXiv  Detail & Related papers  (2023-05-24T15:39:43Z)
• VX2TEXT: End-to-End Learning of Video-Based Text Generation From Multimodal Inputs [103.99315770490163]
  We present a framework for text generation from multimodal inputs consisting of video plus text, speech, or audio. Experiments demonstrate that our approach based on a single architecture outperforms the state-of-the-art on three video-based text-generation tasks.
  arXiv  Detail & Related papers  (2021-01-28T15:22:36Z)
• Bridging the Modality Gap for Speech-to-Text Translation [57.47099674461832]
  End-to-end speech translation aims to translate speech in one language into text in another language via an end-to-end way. Most existing methods employ an encoder-decoder structure with a single encoder to learn acoustic representation and semantic information simultaneously. We propose a Speech-to-Text Adaptation for Speech Translation model which aims to improve the end-to-end model performance by bridging the modality gap between speech and text.
  arXiv  Detail & Related papers  (2020-10-28T12:33:04Z)

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.