A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, et al.
HuBERT uses offline clustering to provide aligned targets for masked prediction, matching or improving wav2vec 2.0 on speech benchmarks.
Lee D. Erman, Frederick Hayes‐Roth, Victor Lesser, et al.
Hearsay-II introduced a blackboard-style framework that coordinates independent knowledge sources to resolve uncertainty in speech understanding, influencing cooperative AI problem-solving.
Unknown
Data2vec is a general self-supervised learning framework that uses the same learning method for speech, NLP, and computer vision.
Chao-Han Huck Yang, Zih-Ching Chen, Piotr Zelasko, et al.
Voice Memory is an inference-only scheme for agentic speech recognition that uses a frozen corrector and score-gated optimizer to reduce over-correction and improve word error rate across domains.
Pete Burnap, Matthew Williams
This paper applies supervised machine learning and statistical modeling to classify and forecast cyber hate speech on Twitter following a real-world event, demonstrating policy applications.
Kate Hone, Robert Graham
This paper reports the first stage in developing a valid, reliable questionnaire (SASSI) for subjectively evaluating speech recognition system interfaces.
Xuedong Huang, Alex Acero, Hsiao-Wuen Hon, et al.
A comprehensive guide to spoken language processing covering theory, algorithms, and system development from speech recognition to synthesis and understanding.
Mingzhi Chen, Taiming Lu, Jiacheng Zhu, et al.
Introduces Derf, a normalization-free pointwise function that outperforms LayerNorm, RMSNorm, and DyT across vision, speech, and DNA tasks.
Unknown
Dynamic Tanh (DyT) replaces normalization layers in Transformers with a simple element-wise operation, achieving comparable or better performance across vision, language, and speech tasks.
Unknown
Llama 3.1 adds multimodal capabilities via compositional cross-attention adapters for vision and speech, achieving competitive performance with GPT-4V and strong video reasoning.
Yi Luo, Nima Mesgarani
Conv-TasNet is a fully-convolutional time-domain audio separation network that outperforms ideal time-frequency magnitude masks for speech separation.
Daniel Povey
Kaldi is a free, open-source speech recognition toolkit based on finite-state transducers, supporting SGMM and GMM acoustic models.