XTREME-S: Evaluating Cross-lingual Speech Representations
원문 초록
We introduce XTREME-S, a new benchmark to evaluate universal cross-lingual speech representations in many languages. XTREME-S covers four task families: speech recognition, classification, speech-to-text translation and retrieval. Covering 102 languages from 10+ language families, 3 different domains and 4 task families, XTREME-S aims to simplify multilingual speech representation evaluation, as well as catalyze research in "universal" speech representation learning. This paper describes the new benchmark and establishes the first speech-only and speech-text baselines using XLS-R and mSLAM on all downstream tasks. We motivate the design choices and detail how to use the benchmark. Datasets and fine-tuning scripts are made easily accessible at https://hf.co/datasets/google/xtreme_s.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
KaLM-Embedding-V2: Superior Training Techniques and Data Inspire A Versatile Embedding Model
Language-Guided Contrastive Audio-Visual Masked Autoencoder with Automatically Generated Audio-Visual-Text Triplets from Videos
Multimodal Representation Learning and Fusion
LM-SPT: LM-Aligned Semantic Distillation for Speech Tokenization
H-QuEST: Accelerating Query-by-Example Spoken Term Detection with Hierarchical Indexing
Transformer-Based Named Entity Recognition for Automated Server Provisioning
토론
이 논문과 연결된 포럼 글이 아직 없습니다.