Title:Samba-ASR: State-Of-The-Art Speech Recognition Leveraging Structured State-Space Models

Abstract:We propose Samba ASR,the first state of the art Automatic Speech Recognition(ASR)model leveraging the novel Mamba architecture as both encoder and decoder,built on the foundation of state space models(SSMs).Unlike transformerbased ASR models,which rely on self-attention mechanisms to capture dependencies,Samba ASR effectively models both local and global temporal dependencies using efficient statespace dynamics,achieving remarkable performance this http URL addressing the limitations of transformers,such as quadratic scaling with input length and difficulty in handling longrange dependencies,Samba ASR achieves superior accuracy and this http URL results demonstrate that Samba ASR surpasses existing opensource transformerbased ASR models across various standard benchmarks,establishing it as the new state of theart in this http URL evaluations on the benchmark dataset show significant improvements in Word Error Rate(WER),with competitive performance even in lowresource this http URL,the inherent computational efficiency and parameter optimization of the Mamba architecture make Samba ASR a scalable and robust solution for diverse ASR this http URL contributions include the development of a new Samba ASR architecture for automatic speech recognition(ASR),demonstrating the superiority of structured statespace models(SSMs)over transformer based models for speech sequence this http URL provide a comprehensive evaluation on public benchmarks,showcasing stateoftheart(SOTA)performance,and present an indepth analysis of computational efficiency,robustness to noise,and sequence this http URL work highlights the viability of Mamba SSMs as a transformerfree alternative for efficient and accurate this http URL leveraging the advancements of statespace modeling,Samba ASR redefines ASR performance standards and sets a new benchmark for future research in this field.