HyperAIHyperAI
StartseitePlattformDokumentationNeuigkeitenForschungsarbeitenTutorialsDatensätzeWikiSOTALLM-ModelleGPU-RanglisteVeranstaltungen
Suche
Über
HyperAIHyperAI

Command Palette

Search for a command to run...

Speech Recognition On Librispeech Test Clean

Metriken

Word Error Rate (WER)

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

Paper TitleRepository
AmNet8.6Amortized Neural Networks for Low-Latency Speech Recognition-
HMM-(SAT)GMM8.0--
Local Prior Matching (Large Model)7.19Semi-Supervised Speech Recognition via Local Prior Matching
Snips6.4Snips Voice Platform: an embedded Spoken Language Understanding system for private-by-design voice interfaces
Li-GRU6.2The PyTorch-Kaldi Speech Recognition Toolkit
HMM-DNN + pNorm*5.5--
CTC + policy learning5.42Improving End-to-End Speech Recognition with Policy Learning-
Deep Speech 25.33Deep Speech 2: End-to-End Speech Recognition in English and Mandarin
Gated ConvNets4.8Letter-Based Speech Recognition with Gated ConvNets
HMM-TDNN + iVectors4.8--
Centaurus (30 M)4.4Let SSMs be ConvNets: State-space Modeling with Optimal Tensor Contractions-
HMM-TDNN trained with MMI + data augmentation (speed) + iVectors + 3 regularizations4.3--
CTC-CRF 4gram-LM4.09CRF-based Single-stage Acoustic Modeling with CTC Topology-
Seq-to-seq attention3.82Improved training of end-to-end attention models for speech recognition
Model Unit Exploration3.60On the Choice of Modeling Unit for Sequence-to-Sequence Speech Recognition
MT4SSL3.4MT4SSL: Boosting Self-Supervised Speech Representation Learning by Integrating Multiple Targets
Convolutional Speech Recognition3.26Fully Convolutional Speech Recognition-
tdnn + chain + rnnlm rescoring3.06Neural Network Language Modeling with Letter-based Features and Importance Sampling-
Jasper DR 10x52.95Jasper: An End-to-End Convolutional Neural Acoustic Model
Jasper DR 10x5 (+ Time/Freq Masks)2.84Jasper: An End-to-End Convolutional Neural Acoustic Model
0 of 64 row(s) selected.