A Comparative Analysis on ASR System Combination for Attention, CTC, Factored Hybrid, and Transducer Models

Documenti analoghi

• Right Label Context in End-to-End Training of Time-Synchronous ASR Models
  di: Raissi, Tina, et al.
  Pubblicazione: (2025)
• Chunked Attention-based Encoder-Decoder Model for Streaming Speech Recognition
  di: Zeineldeen, Mohammad, et al.
  Pubblicazione: (2023)
• The Conformer Encoder May Reverse the Time Dimension
  di: Schmitt, Robin, et al.
  Pubblicazione: (2024)
• Investigating the Effect of Label Topology and Training Criterion on ASR Performance and Alignment Quality
  di: Raissi, Tina, et al.
  Pubblicazione: (2024)
• Analysis of Domain Shift across ASR Architectures via TTS-Enabled Separation of Target Domain and Acoustic Conditions
  di: Raissi, Tina, et al.
  Pubblicazione: (2025)