Large language models (LLMs) have become proficient at solving a wide variety of tasks, including those involving multi-modal inputs. In particular, instantiating an LLM (such as LLaMA) with a speech encoder and training it on paired data imparts speech recognition (ASR) abilities to the decoder-only model, hence called Speech-LLaMA. Nevertheless, due to the sequential nature of auto-regressive inference and the relatively large decoder, Speech-LLaMA models require relatively high inference time. In this work, we propose to speed up Speech-LLaMA inference by predicting multiple tokens in the same decoding step. We explore several model architectures that enable this, and investigate their performance using threshold-based and verification-based inference strategies. We also propose a prefix-based beam search decoding method that allows efficient minimum word error rate (MWER) training for such models. We evaluate our models on a variety of public benchmarks, where they reduce the number of decoder calls by ~3.2x while maintaining or improving WER performance.
Discrete Tokens Exhibit Interlanguage Speech Intelligibility Benefit: an Analytical Study Towards Accent-robust ASR Only with Native Speech Data
The University of Tokyo（东京大学）；Japan National Institute of Advanced Industrial Science and Technology (AIST)（日本国家工业科学与技术研究院）；Japan Kyoto University（京都大学）
Token-level Speaker Change Detection Using Speaker Difference and Speech Content via Continuous Integrate-and-fire
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Bytedance AI LAB（字节跳动人工智能实验室）
Prosodically Enhanced Foreign Accent Simulation by Discrete Token-based Resynthesis Only with Native Speech Corpora
The University of Tokyo（东京大学）；Japan National Institute of Advanced Industrial Science and Technology (AIST)（日本国家先进工业科学与技术研究院）；Japan Kyoto University（京都大学）
Whispering LLaMA: A Cross-Modal Generative Error Correction Framework for Speech Recognition
King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）；Manipal Institute of Technology（马尼帕尔理工学院）；Georgia Institute of Technology（佐治亚理工学院）；Karolinska Institute（卡罗林斯卡学院）；SDAIA-KAUST Center of Excellence in Data Science and Artificial Intelligence（SDAIA-KAUST数据科学与人工智能卓越中心）
TokAN: Accent Normalization Using Self-Supervised Speech Tokens
School of Data Science (SDS), The Chinese University of Hong Kong, Shenzhen (CUHKSZ)（香港中文大学（深圳）数据科学学院）；Tencent Ethereal Audio Lab, Tencent（腾讯虚幻音频实验室）；School of Intelligence Science and Technology, Nanjing University, Suzhou（南京大学苏州校区智能科学与技术学院）；Shenzhen Loop Area Institute（深圳河套学院）；X-LANCE Lab, School of Computer Science, Shanghai Jiao Tong University（上海交通大学计算机科学与工程学院X-LANCE实验室）；School of Artificial Intelligence (SAI), CUHKSZ（香港中文大学（深圳）人工智能学院）；Shenzhen Research Institute of Big Data（深圳大数据研究院）
DM-Codec: Distilling Multimodal Representations for Speech Tokenization
Center for Computational & Data Sciences, Independent University, Bangladesh（计算与数据科学中心，独立大学，孟加拉国）；Amazon GenAI（亚马逊生成人工智能）；Qatar Computing Research Institute（卡塔尔计算研究所）；University of Virginia（弗吉尼亚大学）
Advanced Modeling of Interlanguage Speech Intelligibility Benefit with L1-L2 Multi-Task Learning Using Differentiable K-Means for Accent-Robust Discrete Token-Based ASR
The University of Tokyo（东京大学）；National Institute of Advanced Industrial Science and Technology (AIST)（国家工业科学与技术研究院）
Contextualized Token Discrimination for Speech Search Query Correction
AI Group, WeBank（WeBank人工智能组）；Hong Kong Polytechnic University（香港理工大学）；Macau University of Science and Technology（澳门科学理工学院）；Hong Kong University of Science and Technology（香港理工大学）
LLaMA-Omni: Seamless Speech Interaction with Large Language Models
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；Key Laboratory of AI Safety, Chinese Academy of Sciences（中国科学院人工智能安全重点实验室）；University of Chinese Academy of Sciences（中国科学院大学）
LLaMA-Omni2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech Synthesis
Key Laboratory of Intelligent Information Processing Institute of Computing Technology, Chinese Academy of Sciences (ICT/CAS)（智能信息处理重点实验室，计算技术研究所，中国科学院）；Key Laboratory of AI Safety, Chinese Academy of Sciences（人工智能安全重点实验室，中国科学院）；University of Chinese Academy of Sciences, Beijing, China（中国科学院大学，北京，中国）
Acoustic BPE for Speech Generation with Discrete Tokens
MoE Key Lab of Artificial Intelligence, AI Institute（教育部人工智能重点实验室，人工智能研究院）；X-LANCE Lab, Department of Computer Science and Engineering（X-LANCE实验室，计算机科学与工程系）；Shanghai Jiao Tong University（上海交通大学）
Fewer-token Neural Speech Codec with Time-invariant Codes
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Tsinghua University（清华大学）
UniCATS: A Unified Context-Aware Text-to-Speech Framework with Contextual VQ-Diffusion and Vocoding
MoE Key Lab of Artificial Intelligence, AI Institute X-LANCE Lab, Department of Computer Science and Engineering, Shanghai Jiao Tong University（上海交通大学电子信息与电气工程学院计算机科学与工程系人工智能教育部重点实验室X-LANCE实验室）；Shenzhen Research Institute of Big Data（深圳市大数据研究院）；AISpeech Ltd（思必驰科技股份有限公司）
