Speech Recognition

Spectrograms, CTC, RNN-T, Conformer, Whisper, streaming, diarisation and self-supervised audio.

20concepts
140flashcards
156minutes of reading
  1. 01 Beam Search Decoding in ASR Beam search decoding navigates the exponentially large label sequence space in ASR by maintaining a fixed-width frontier of the most probable partial hypotheses at each step, making it the default inference strategy for CTC, RNN-T, and attention-based models. intermediate 8m 7 cards
  2. 02 CTC: Connectionist Temporal Classification CTC is a training objective that lets a neural network learn to align variable-length audio to text without any hand-labelled frame-level annotations. intermediate 8m 7 cards
  3. 03 Causal and Chunked Attention for Streaming Streaming ASR requires attention mechanisms that never look at future audio; causal masking and chunked attention are the two principal techniques, each trading latency against accuracy in different ways. intermediate 8m 7 cards
  4. 04 Endpointing and Voice Activity Detection Endpointing and voice activity detection are the mechanisms that decide when a user has finished speaking, directly controlling the latency and correctness of every streaming ASR system. intermediate 8m 7 cards
  5. 05 Listen, Attend and Spell LAS is a purely sequence-to-sequence ASR model that replaces HMMs, CTC, and explicit pronunciation dictionaries with a pyramidal RNN encoder and an attention-based character decoder trained end-to-end. intermediate 8m 7 cards
  6. 06 Robustness to Noise and Accents How modern ASR systems are trained and adapted to handle environmental noise, channel distortions, and speaker accent variability without collapsing to near-zero accuracy. intermediate 8m 7 cards
  7. 07 Speaker Diarisation Speaker diarisation segments an audio recording into speaker-homogeneous regions and assigns each region a speaker identity, answering the question "who spoke when" without necessarily transcribing what was said. intermediate 8m 7 cards
  8. 08 Streaming vs Offline ASR Streaming ASR emits partial transcripts incrementally as audio arrives, trading access to future context for low latency, while offline ASR processes the full utterance and consistently achieves lower word error rates. intermediate 7m 7 cards
  9. 09 The CTC Blank Token and Alignment The CTC blank token is a special output symbol that lets a neural network emit one label per time-step without needing a hand-crafted alignment between audio frames and characters. intermediate 7m 7 cards
  10. 10 The Conformer Architecture The Conformer interleaves convolution and multi-head self-attention inside each encoder block to capture both fine-grained local acoustic patterns and long-range sequence dependencies, achieving state-of-the-art ASR accuracy on LibriSpeech. intermediate 8m 7 cards
  11. 11 Whisper's Multitask Decoder Whisper conditions a single sequence-to-sequence decoder on a prefix of special tokens that specify language, task, and timestamp behaviour, allowing one model to handle transcription, translation, and language identification without any task-specific heads. intermediate 7m 7 cards
  12. 12 Whisper: Weakly-Supervised ASR Whisper trains a sequence-to-sequence Transformer on 680,000 hours of weakly-supervised internet audio to achieve robust multilingual speech recognition without task-specific fine-tuning. intermediate 8m 7 cards
  13. 13 Why Sequence Length Is Hard in Audio Audio produces roughly 100 frames per second of speech, making sequence lengths 10-50x longer than equivalent text, and this mismatch drives almost every architectural and training decision in modern ASR. intermediate 7m 7 cards
  14. 14 Word Error Rate and ASR Evaluation Word Error Rate measures ASR accuracy as the minimum edit distance between a hypothesis and a reference transcript, normalised by reference length, but its apparent simplicity hides a nest of normalisation choices that make numbers across papers routinely incomparable. intermediate 8m 7 cards
  15. 15 wav2vec 2.0 and Self-Supervised Audio wav2vec 2.0 learns speech representations from raw audio without transcripts by masking latent features and solving a contrastive task over learned discrete units, then fine-tunes on as little as ten minutes of labelled speech to reach competitive word error rates. intermediate 8m 7 cards