Speech recognition: A model and a program for research

By Morris Halle, Kenneth N. Stevens. 1962

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Dong Yu
2. Li Deng
3. Wei-Ning Hsu
4. Jinyu Li
5. Yoshua Bengio
6. Abdel-rahman Mohamed
7. James R. Glass
8. Yu Zhang

Citations in the corpus

• Does English coronal place assimilation create lexical ambiguity (2002, background): “…articulation in perception argue that listeners recognize speech through a series of mappings that trace the speech chain backward from perception through production to the speaker’s initial intent (Halle & Stevens, 1962; Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967; Mann & Repp, 1980).”
• The motor theory of speech perception reviewed (2006, background): “They proposed analysis by synthesis (cf. Halle & Stevens, 1962; Stevens, 1960), in which the listener analyzed the acoustic input, guessed at how it was produced by the speaker, synthesized a virtual acoustic signal based on the guess, and matched the virtual to the actual signal.”
• A Convolutional Deep Markov Model for Unsupervised Speech Representation Learning (2020, methodology): “The computational modeling framework of generative models is well-suited for the problem of spoken language acquisition, as it relates to the classic analysis-by-synthesis theories of speech recognition [6, 7].”
• Discovering linguistic structures in speech: models and applications (2014, background): “Can the same approach be applied to design a computational model that achieves human-level performance on one-shot learning tasks for spoken words? More specifically, the term of analysis-by-synthesis in the speech community often refers to explicit modeling of the articulatory process [65], which r”
• Binary coding of speech spectrograms using a deep auto-encoder (2010, background): “…of generative modeling and of exploiting the multiple layers of non-linear transformations in the human speech generation process (e.g., [1][2][3][4][5][6]) it was not until the introduction of generative, layer-by-layer pre-training that the full power of generative modeling became apparent.”
• Unified Coding of Spectral and Temporal Phonetic Cues: Electrophysiological Evidence for Abstract Phonological Features (2021, background): “…the current results indicate a brain signature underlying abstract groupings of speech sounds and are consistent with predictive theories of processing that are at least partially based on phonological features, such as analysis by synthesis (Halle & Stevens, 1962; Poeppel & Monahan, 2011).”