IP Library Granted Patent US 12,737,640
Granted Patent B2
US 12,737,640 · App. 17/899,913 · Granted Sep 15, 2026

Controllable dynamic multi-task architectures

Inventors: Yumin Suh (Santa Clara, CA); Samuel Schulter (New York, NY); Xiang Yu (Mountain View, CA); Masoud Faraki (San Francisco, CA); Manmohan Chandraker (Santa Clara, CA); Dripta Raychaudhuri (Riverside, CA)
Assignee: NEC Corporation
G06N3/0985
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,737,640
App. No.
17/899,913
Granted
Sep 15, 2026
Kind
B2
Abstract

Systems and methods for generating a hypernetwork configured to be trained for a plurality of tasks; receiving a task preference vector identifying a hierarchical priority for the plurality of tasks, and a resource constraint as a tuple; finding tree sub-structures and the corresponding modulation of features for every tuple within an N-stream anchor network; optimizing a branching regularized loss function to train an edge hypernet; and training a weight hypernet, keeping the anchor net and the edge hypernet fixed.

Claims (225)

1 . A method for building a dynamic multi-task network, comprising:

generating a hypernetwork configured to model tradeoffs among a plurality of competing tasks;

receiving a task preference vector identifying a hierarchical priority for the plurality of competing tasks, and a resource constraint as a tuple;

finding tree sub-structures and the corresponding modulation of features for every tuple within an N-stream anchor network;

optimizing a branching regularized loss function to train an edge hypernet, wherein the branching regularized loss function is based on an active loss L active and an inactive loss, the active loss, L active , is calculated as:

active

=

l

=

1

L

i

,

j

𝒜

i

j

L

-

l

L

·

A

(

i

,

j

)

·

P

(

l

,

i

,

j

)

·

v

i

l

-

v

j

l

2

where P(l, i, j)=P use (l, i)>P use (l, j) is the probability that the nodes i and j in layer l are used in the sampled tree structure, and A(i, j) captures the task affinity between tasks, τ i and τ j , and

υ

i

l

and

υ

j

l

are samples, represents active tasks and L represents the total number of layers;

training a weight hypernet, keeping the anchor network and the edge hypernet fixed; and

generating, by the hypernetwork before inference time, an architectural configuration for a main model based on the task preference vector and the resource constraint.

2 . The method as recited in claim 1 , wherein the N-stream anchor network has fixed weights for finding the tree sub-structures.

3 . The method as recited in claim 1 , wherein finding the tree sub-structures includes selecting a parent from every node.

4 . The method as recited in claim 1 , wherein the edge hypernet predicts the branching parameters within the anchor network.

5 . The method as recited in claim 1 , further comprising optimizing a task loss, L task , by taking into account the individual competing task performances without considering a computational cost.

6 . The method as recited in claim 1 , wherein competing tasks with higher preferences have a greater influence.

7 . The method as recited in claim 1 , wherein the active loss is additionally weighted by a cost preference, c, of the resource constraint tuple to enable the control of total computational cost.

8 . A computer system for building a dynamic multi-task network, comprising:

one or more processors;

a memory in communication with the one or more processors; and

a dynamic multi-task network builder stored in the memory, wherein the dynamic multi-task network builder performs;

generate a hypernetwork configured to model tradeoffs among a plurality of competing tasks;

receive a task preference vector identifying a hierarchical priority for the plurality of competing tasks, and a resource constraint as a tuple;

find tree sub-structures and the corresponding modulation of features for every tuple within an N-stream anchor network;

optimize a branching regularized loss function to train an edge hypernet, wherein the branching regularized loss function is based on an active loss L active and an inactive loss, the active loss, L active , is calculated as:

active

=

l

=

1

L

i

,

j

𝒜

i

j

L

-

l

L

·

A

(

i

,

j

)

·

P

(

l

,

i

,

j

)

·

v

i

l

-

v

j

l

2

where P(l, i, j)=P use (l, i)·P use (l, j) is the probability that the nodes i and j in layer l are used in the sampled tree structure, and A(i, j) captures the task affinity between tasks, τ i and τ j , and

υ

i

l

and

υ

j

l

are samples, represents active tasks and L represents the total number of layers;

training a weight hypernet, keeping the anchor net and the edge hypernet fixed; and

generating, by the hypernetwork before inference time, an architectural configuration for a main model based on the task preference vector and the resource constraint.

9 . The computer system for building a dynamic multi-task network of claim 8 , wherein the N-stream anchor network has fixed weights for finding the tree sub-structures.

10 . The computer system for building a dynamic multi-task network of claim 8 , wherein finding the tree sub-structures includes selecting a parent from every node.

11 . The computer system for building a dynamic multi-task network of claim 8 , wherein the edge hypernet predicts the branching parameters within the anchor network.

12 . The computer system for building a dynamic multi-task network of claim 8 , further comprising optimizing a task loss, L task , by taking into account the individual competing task performances without considering a computational cost.

13 . The computer system for building a dynamic multi-task network of claim 8 , wherein competing tasks with higher preferences have a greater influence.

14 . The computer system for building a dynamic multi-task network of claim 8 , wherein the active loss is additionally weighted by a cost preference, c, of the resource constraint tuple to enable the control of total computational cost.

15 . A non-transitory computer program product for building a dynamic multi-task network, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform a method comprising:

generating a hypernetwork configured to model tradeoffs among a plurality of competing tasks;

receiving a task preference vector identifying a hierarchical priority for the plurality of competing tasks, and a resource constraint as a tuple;

finding tree sub-structures and the corresponding modulation of features for every tuple within an N-stream anchor network;

optimizing a branching regularized loss function to train an edge hypernet, wherein the branching regularized loss function is based on an active loss L active and an inactive loss, the active loss, L active , is calculated as:

active

=

l

=

1

L

i

,

j

𝒜

i

j

L

-

l

L

·

A

(

i

,

j

)

·

P

(

l

,

i

,

j

)

·

v

i

l

-

v

j

l

2

where P(l, i, j)=P use (l, i)·P use (l,j) is the probability that the nodes i and j in layer l are used in the sampled tree structure, and A(i,j) captures the task affinity between tasks, τ i and τ j , and

υ

i

l

and

υ

j

l

are samples, represents active tasks and L represents the total number of layers;

calculating an active loss and an inactive loss to determine branching within the anchor network

training a weight hypernet, keeping the anchor network and the edge hypernet fixed; and

generating, by the hypernetwork before inference time, an architectural configuration for a main model based on the task preference vector and the resource constraint.

16 . The computer program product as recited in claim 15 , wherein the N-stream anchor network has fixed weights for finding the tree sub-structures.

17 . The computer program product as recited in claim 15 , wherein finding the tree sub-structures includes selecting a parent from every node.

18 . The computer program product as recited in claim 15 , wherein the edge hypernet predicts the branching parameters within the anchor network.

19 . The computer program product as recited in claim 15 , further comprising optimizing a task loss, L task , by taking into account the individual competing task performances without considering a computational cost.

20 . The computer program product as recited in claim 15 , wherein competing tasks with higher preferences have a greater influence.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 14, 2026
From: NEC LABORATORIES AMERICA, INC.
To: NEC CORPORATION
Reel/Frame 075265/0791 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 31, 2022
From: SUH, YUMIN; SCHULTER, SAMUEL; YU, XIANG; FARAKI, MASOUD; CHANDRAKER, MANMOHAN; RAYCHAUDHURI, DRIPTA
To: NEC LABORATORIES AMERICA, INC.
Reel/Frame 060952/0941 →
Continuity (3)
Provisional Application 63279717 · Nov 16, 2021
Provisional Application 63240522 · Sep 3, 2021
Related Publication 20230196122A1 · Jun 22, 2023
References Cited (13)
US 20190130257A1 · Meyerson · 2019 [cited by examiner]
US 20230076457A1 · Gao · 2023 [cited by examiner]
Bruggemann et al., “Automated Search for Resource-Efficient Branched Multi-Task Networks” (Year: 2020). [cited by examiner]
Guo et al., “Learning to Branch for Multi-task Learning” (Year: 2020). [cited by examiner]
Lin et al., “Controllable Pareto Multi-task Learning” (Year: 2021). [cited by examiner]
Mahabadi et al., “Parameter-efficient Multi-task Fine-tuning for Transformers via Shared Hypernetworks” (Year: 2021). [cited by examiner]
Sarafian et al., “Recomposing the Reinforcement Learning Building Blocks with Hypernetworks” (Year: 2021). [cited by examiner]
Liu, H., Simonyan, K., & Yang, Y. (Jun. 24, 2018). Darts: Differentiable architecture search. arXiv preprint arXiv:1806.09055. [cited by applicant]
Navon, A., Shamsian, A., Chechik, G., & Fetaya, E. (Oct. 8, 2020). Learning the pareto front with hypernetworks. arXiv preprint arXiv:2010.04104. [cited by applicant]
Lin, X., Yang, Z., Zhang, Q., & Kwong, S. (Oct. 13, 2020). Controllable pareto multi-task learning. arXiv preprint arXiv:2010.06313. [cited by applicant]
Wang, X., Yu, F., Dou, Z. Y., Darrell, T., & Gonzalez, J. E. (Sep. 8, 2018). Skipnet: Learning dynamic routing in convolutional networks. In Proceedings of the European Conference on Computer Vision (ECCV) (pp. 409-424). [cited by applicant]
Dosovitskiy, A., & Djolonga, J. (Sep. 25, 2019). You only train once: Loss-conditional training of deep networks. In International conference on learning representations. [cited by applicant]
Yuan, Z., Wu, B., Sun, G., Liang, Z., Zhao, S., & Bi, W. (Aug. 23, 2020). S2dnas: Transforming static cnn model for dynamic inference via neural architecture search. In European Conference on Computer Vision (pp. 175-19… [cited by applicant]