View source for Google llc (20250131273). Learning the Joint Distribution of Two Sequences Using Little or No Paired Data

=Learning the Joint Distribution of Two Sequences Using Little or No Paired Data=


==Organization Name==

[[:Category:google llc|google llc]]


[[Category:google llc]]

==Inventor(s)==

[[:Category:Soroosh Mariooryad of Redwood City CA US|Soroosh Mariooryad of Redwood City CA US]][[Category:Soroosh Mariooryad of Redwood City CA US]]

[[:Category:Sean Matthew Shannon of San Francisco CA US|Sean Matthew Shannon of San Francisco CA US]][[Category:Sean Matthew Shannon of San Francisco CA US]]

[[:Category:Thomas Edward Bagby of Monte Rio CA US|Thomas Edward Bagby of Monte Rio CA US]][[Category:Thomas Edward Bagby of Monte Rio CA US]]

[[:Category:Siyuan Ma of San Jose CA US|Siyuan Ma of San Jose CA US]][[Category:Siyuan Ma of San Jose CA US]]

[[:Category:David Teh-Hwa Kao of Philadelphia PA US|David Teh-Hwa Kao of Philadelphia PA US]][[Category:David Teh-Hwa Kao of Philadelphia PA US]]

[[:Category:Daisy Antonia Stanton of San Francisco CA US|Daisy Antonia Stanton of San Francisco CA US]][[Category:Daisy Antonia Stanton of San Francisco CA US]]

[[:Category:Eric Dean Battenberg of Walnut Creek CA US|Eric Dean Battenberg of Walnut Creek CA US]][[Category:Eric Dean Battenberg of Walnut Creek CA US]]

[[:Category:Russell John Wyatt Skerry-ryan of Mountain View CA US|Russell John Wyatt Skerry-ryan of Mountain View CA US]][[Category:Russell John Wyatt Skerry-ryan of Mountain View CA US]]

==Learning the Joint Distribution of Two Sequences Using Little or No Paired Data==


This abstract first appeared for US patent application 20250131273 titled 'Learning the Joint Distribution of Two Sequences Using Little or No Paired Data

==Original Abstract Submitted==

provided is a noisy channel generative model of two sequences, for example text and speech, which enables uncovering the associations between the two modalities when limited paired data is available. to address the intractability of the exact model under a realistic data set-up, example aspects of the present disclosure include a variational inference approximation. to train this variational model with categorical data, a kl encoder loss approach is proposed which has connections to the wake-sleep algorithm.



[[Category:G06N3/088]]
[[Category:G06N3/0455]]
[[Category:CPC_G06N3/088]]