Janet M. Baker

Janet M. Baker
Alma mater	Tufts University Carnegie Mellon University
Known for	Dragon Systems
Spouse	James K. Baker
Scientific career
Fields	speech recognition
Institutions	MIT Media Lab Harvard Medical School Dragon Systems
Thesis	A new time-domain analysis of human speech and other complex waveforms  (1975)
Doctoral advisor	Raj Reddy

Janet MacIver Baker is an American computer scientist, neuroscientist and entrepreneur. Along with her husband James K. Baker, they founded Dragon Systems and are together credited with the creation of Dragon NaturallySpeaking. ^[1]

In 2012, she received the IEEE James L. Flanagan Speech and Audio Processing Award with her husband. ^[2]

Baker is currently affiliated with the MIT Media Lab and Harvard Medical School as a visiting scientist and lecturer. ^[3]

Early life and education

Baker trained as a biophysicist and pursued graduate studies at Rockefeller University in New York City beginning in 1970. ^[4]

Baker transferred with James Baker to Carnegie Mellon University, a major center for artificial intelligence and speech understanding research. There, she completed her doctoral training and collaborated on foundational statistical approaches to speech recognition. ^[5]

Research career

At Carnegie Mellon in the early 1970s, Baker worked on speech recognition at a time when the dominant paradigm emphasized rule-based linguistics and symbolic artificial intelligence. ^[6] In contrast, she and James Baker pursued a statistical, data-driven approach, applying probabilistic modeling techniques later known as Hidden Markov models (HMMs) to continuous speech. ^[7]

After earning her doctorate in 1975, Baker joined IBM’s Thomas J. Watson Research Center, where she continued research on large-vocabulary, continuous speech recognition. ^[8]

Dragon Systems

In 1982, Baker co-founded Dragon Systems with James Baker after the couple left IBM and later Exxon subsidiary Verbex. ^[9] She served as president of the company, while James Baker served as chairman and CEO. Dragon Systems initially operated without venture capital, relying on revenue from contracts and early speech-recognition products. ^[10]

In 1997, Dragon introduced Dragon NaturallySpeaking, the first continuous speech recognition software for desktop computers. ^[11]

References

1. "History of Speech Recognition". Dragon Medical Transcription. Archived from the original on August 13, 2015. Retrieved January 17, 2015.
2. "IEEE James L. Flanagan Speech and Audio Processing Award Recipients". IEEE. Archived from the original on April 7, 2010. Retrieved July 2, 2017.
3. Blanding, Michael (Fall 2012). "Speechless". Tufts Magazine. Archived from the original on May 1, 2016. Retrieved July 2, 2017.
4. "Teaching a Computer to Hear | Tufts Now". now.tufts.edu. December 4, 2012. Retrieved December 15, 2025.
5. Simson Garfinkel ’87, PhD ’05archive page. "Enter the Dragon". MIT Technology Review. Retrieved December 15, 2025.
6. "Milestones in Speech Technology – Past and Future!". Speech Technology Magazine. August 30, 2005. Retrieved December 15, 2025.
7. "First-Hand:The Hidden Markov Model". ETHW. January 12, 2015. Retrieved December 15, 2025.
8. Simson Garfinkel ’87, PhD ’05archive page. "Enter the Dragon". MIT Technology Review. Retrieved December 15, 2025.
9. "Goldman Sachs and the $580 Million Black Hole (Published 2012)". July 14, 2012. Archived from the original on July 8, 2023. Retrieved December 15, 2025.
10. Baker, J. (February 1975). "The DRAGON system–An overview". IEEE Transactions on Acoustics, Speech, and Signal Processing. 23 (1): 24–29. doi:10.1109/TASSP.1975.1162650. ISSN   0096-3518.
11. "Nuance Dragon Naturally Speaking, Medical Transcription, Voice Recognition Software". www.dragon-medical-transcription.com. Retrieved December 15, 2025.