# US 2020/0286471 A1 · System to convert phonemes into phonetics-based words

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0286471 A1
- Document ID: US20200286471A1
- Status: Published application
- Kind: A1
- Application number: 16/810,373
- Filed: 2020-03-05
- Published: 2020-09-10
- Earliest claimed: 2019-03-05
- Assignees: Medyug Technology Private Limited
- Applicants: Medyug Technology Private Limited
- Inventors: Baljit Singh; Praveen Prakash
- Web page: https://patentgrep.com/patent/US20200286471A1

## CPC

- G10L 15/187: Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
- G10L 15/02: Feature extraction for speech recognition; Selection of recognition unit
- G10L 15/22: Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L 15/24: Speech recognition using non-acoustical features
- G10L 15/197: Probabilistic grammars, e.g. word n-grams
- G10L 2015/025: {Phonemes, fenemes or fenones being the recognition units}
- G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING

## Abstract

A system to convert phonemes into phonetics-based words that is implemented in one or more computing systems, in association with a system that provides required inputs is disclosed. Said system comprises a phoneme enhancer, a phoneme sequence buffer, a phoneme sequence to phonetics-based word converter that comprises a sliding window phoneme sequence matcher, a phoneme sequence to phonetics-based word custom data memory, a most frequent phonetics-based word predictive memory, a phoneme similarity matrix, and a phonetics-based word output unit.

## Claims

### Claim 1

A system to convert phonemes into phonetics-based words that is implemented in one or more computing systems in association with a system that provides required inputs, comprising:
a phoneme enhancer that takes both voice sounds and visual hints as inputs, and merges them to form a phoneme;
a phoneme sequence buffer that facilitates the storing of an incoming sequence of phonemes, until it is clustered into a meaningful thought representation;
a phoneme sequence to phonetics-based word custom data memory that takes a sequence of phonemes as input, and outputs: either a fully matching phonetics-based word, or a most matching phonetics-based word;
a most frequent phonetics-based word predictive memory that is configured to imitate a human brain, which can predict what could come next;
a phoneme similarity matrix that is a key value store, which has a list of phonemes that are similar;
a phoneme sequence to phonetics-based word converter that controls all the data flow across various other modules and does reasoning, prioritization, and disambiguation, said phoneme sequence to phonetics-based word converter taking phoneme sequences as input from the phoneme sequence to phonetics-based word custom data memory, the phonetics-based word predictive memory, and the phoneme similarity matrix, to get the best possible phonetics-based word sequence, with: the phoneme sequence to phonetics-based word converter interacting with a master controller of a language understanding system to send and receive signals, said output being forwarded for checking the meaning; and
a phonetics-based word output unit that comprises the sequence of phonetics-based words that are output from the system.

### Claim 2 (depends on 1)

The system to convert phonemes into phonetics-based words as claimed in claim 1, wherein the phoneme sequence to phonetics-based word custom data memory is an associative memory, which is either a look-up memory or is implemented using any statistical Artificial Intelligence modeling.

### Claim 3 (depends on 1)

The system to convert phonemes into phonetics-based words as claimed in claim 1, wherein the system uses International Phonetic Alphabets phonetic standards to represent phonetics-based representation of words.
