IP Library › Granted Patent US 10,886,943
Granted Patent B2
US 10,886,943 · App. 16/576,166 · Granted Jan 5, 2021

Method and apparatus for variable rate compression with a conditional autoencoder

Inventors: Yoo Jin Choi (San Diego, CA); Mostafa El-Khamy (San Diego, CA); Jungwon Lee (San Diego, CA)
H03M7/6011G06N3/04G06N3/08H04L1/0003H04L1/0009H04N19/13H04N19/147H04N19/19
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 10,886,943
App. No.
16/576,166
Filed
Sep 19, 2019
Granted
Jan 5, 2021
Kind
B2
Art Unit
2845
USPC
341/87
Abstract

A method and apparatus for variable rate compression with a conditional autoencoder is herein provided. According to one embodiment, a method includes training a conditional autoencoder using a Lagrange multiplier and training a neural network that includes the conditional autoencoder with mixed quantization bin sizes.

Claims (26)

1. A method for compression, comprising:

training a conditional autoencoder using a Lagrange multiplier; and

training a neural network that includes the conditional autoencoder with mixed quantization bin sizes.

2. The method of claim 1 , wherein a value of the Lagrange multiplier is selected from a predetermined set of values.

3. The method of claim 1 , wherein the neural network produces a compressed out with varying compression rate and distortion depending on the conditioning value of the Lagrange multiplier.

4. The method of claim 1 , wherein the conditional autoencoder is trained using conditional convolution.

5. The method of claim 1 , wherein training the network with mixed quantization bin sizes further comprises adjusting the quantization bin size.

6. The method of claim 1 , wherein training the network with mixed quantization bin sizes further comprises drawing a uniform noise for various noise levels for random quantization bin sizes.

7. The method of claim 1 , wherein the neural network is configured to perform better portable graphics (BPG) compression.

8. The method of claim 1 , wherein the neural network is configured to perform compression of an analog-to-digital output of a radio frequency integrated circuit (RFIC).

9. The method of claim 8 , further comprising selecting a Lagrange multiplier and the quantization bin size as a function of a modulation coding scheme (MCS).

10. The method of claim 9 , wherein the neural network further includes a lookup table that maps the MCS to the Lagrange multiplier and the quantization bin size.

11. A system, comprising:

a neural network including a conditional autoencoder; and

a processor configured to:

train the conditional autoencoder using a Lagrange multiplier; and

train the neural network using mixed quantization bin sizes.

12. The system of claim 11 , wherein a value of the Lagrange multiplier is selected from a predetermined set of values.

13. The system of claim 11 , wherein the neural network produces a compressed out with varying compression rate and distortion depending on the conditioning value of the Lagrange multiplier.

14. The system of claim 11 , wherein the processor is configured to train the conditional autoencoder using conditional convolution.

15. The system of claim 11 , wherein the processor is further configured to train the network with mixed quantization bin sizes by adjusting the quantization bin size.

16. The system of claim 11 , wherein the processor is further configured to train the network with mixed quantization bin sizes by drawing a uniform noise for various noise levels for random quantization bin sizes.

17. The system of claim 11 , wherein the neural network is configured to perform better portable graphics (BPG) compression.

18. The system of claim 11 , wherein the neural network is configured to perform compression of an analog-to-digital converter output of a radio frequency integrated circuit (RFIC).

19. The system of claim 18 , wherein the processor is further configured to select a Lagrange multiplier and the quantization bin size as a function of a modulation coding scheme (MCS).

20. The system of claim 19 , wherein the neural network further includes a lookup table that maps the MCS to the Lagrange multiplier and the quantization bin size.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 8, 2019
From: CHOI, YOO JIN; EL-KHAMY, MOSTAFA; LEE, JUNGWON
To: SAMSUNG ELECTRONICS CO., LTD.
Reel/Frame 050652/0637 →
Continuity (3)
Provisional Application 62860522 · Jun 12, 2019
Provisional Application 62820148 · Mar 18, 2019
Related Publication 20200304147A1 · Sep 24, 2020
Cited By (2)
US 12,256,075 US 12,323,593