IP Library › Granted Patent US 12,417,557
Granted Patent B2
US 12,417,557 · App. 18/473,944 · Granted Sep 16, 2025

Method, apparatus, and storage medium using padding/trimming in compression neural network

Inventors: Joo-Young Lee (Daejeon, KR); Se-Yoon Jeong (Daejeon, KR); Hyoung-Jin Kwon (Daejeon, KR); Dong-Hyun Kim (Daejeon, KR); Youn-Hee Kim (Daejeon, KR); Jong-Ho Kim (Daejeon, KR); Tae-Jin Lee (Daejeon, KR); Jin-Soo Choi (Daejeon, KR)
Assignee: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE
G06T9/002G06T3/4046G06V10/454G06V10/764G06V10/82
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 12,417,557
App. No.
18/473,944
Granted
Sep 16, 2025
Kind
B2
Abstract

An encoding apparatus extracts features of an image by applying multiple padding operations and multiple downscaling operations to an image represented by data and transmits feature information indicating the features to a decoding apparatus. The multiple padding operations and the multiple downscaling operations are applied to the image in an order in which one padding operation is applied and thereafter one downscaling operation corresponding to the padding operation is applied. A decoding method receives feature information from an encoding apparatus, and generates a reconstructed image by applying multiple upscaling operations and multiple trimming operations to an image represented by the feature information. The multiple upscaling operations and the multiple trimming operations are applied to the image in an order in which one upscaling operation is applied and thereafter one trimming operation corresponding to the upscaling operation is applied.

Claims (40)

1. A decoding method, comprising:

receiving image information representing an image; and

generating a decoded image by applying multiple upscaling operations and multiple trimming operations to an image represented by the feature information, wherein

a size of the image is enlarged by n times by each upscaling operation of the multiple upscaling operations,

the size of the image is adjusted to a multiple of n by each trimming operation of the multiple trimming operations,

n is an integer greater than or equal to 2,

the multiple upscaling operations and the multiple trimming operations applied to the image as multiple processes are applied to the image sequentially,

each process of the multiple processes comprises an upscaling operation and a trimming operation on an output from the upscaling operation, and

in a first process and a second process which are adjacent to each other among the multiple processes, an input to a trimming operation of the first process is configured based on an output from an upscaling operation of the first process, and an input to an upscaling operation of the second process is configured based on an output from a trimming operation of the first process.

2. The decoding method of claim 1 , wherein the multiple upscaling operations and the multiple trimming operations are applied to the image in an order in which one upscaling operation is applied to the image and thereafter one trimming operation corresponding to the upscaling operation is applied to the image.

3. The decoding method of claim 1 , wherein number-of-lines information indicating numbers of one or more lines that are added through multiple padding operations performed by an encoding apparatus is received from the encoding apparatus.

4. The decoding method of claim 1 , wherein:

ratio information is received from an encoding apparatus, and

the ratio information includes ratios of multiple downscaling operations performed by the encoding apparatus or reciprocals of the ratios of the multiple downscaling operations performed by the encoding apparatus.

5. The decoding method of claim 4 , wherein:

a k-th upscaling operation, among the multiple upscaling operations, is configured to adjust a size of the image to S k times,

k is an integer that is equal or greater than 1 and less than or equal to m,

m is a number of the multiple downscaling operations, and

S k is determined based on the ratio information.

6. The decoding method of claim 1 , wherein:

a k-th trimming operation on the image, among the multiple trimming operations, is configured to remove, from the image, a number of lines identical to a number of lines that are added through a k-th padding operation corresponding to the k-th trimming operation, among multiple padding operations performed by an encoding apparatus,

k is an integer that is equal to or greater than 1 and less than or equal to m, and

m is a number of the multiple downscaling operations.

7. The decoding method of claim 6 , wherein:

a padding operation performed in a k-th order, among m padding operations performed by the encoding apparatus, corresponds to a trimming operation performed in an (m−k+1)-th order, among m trimming operations on a decoding apparatus performing the decoding method,

k is an integer that is equal to or greater than 1 and less than or equal to m, and

m is a number of the multiple downscaling operations.

8. A non-transitory computer-readable storage medium storing a bitstream, the bitstream comprising:

image information, wherein

a decoded image is generated by applying multiple upscaling operations and multiple trimming operations to an image represented by the image information, wherein

a size of the image is enlarged by n times by each upscaling operation of the multiple upscaling operations,

the size of the image is adjusted to a multiple of n by each trimming operation of the multiple trimming operations,

n is an integer greater than or equal to 2,

the multiple upscaling operations and the multiple trimming operations applied to the image as multiple processes are applied to the image sequentially,

each process of the multiple processes comprises an upscaling operation and a trimming operation on an output from the upscaling operation, and

in a first process and a second process which are adjacent to each other among the multiple processes, an input to a trimming operation the first process is configured based on an output from an upscaling operation of the first process, and an input to an upscaling operation of the second process is configured based on an output from a trimming operation of the first process.

9. The non-transitory computer-readable storage medium of claim 8 , wherein the multiple upscaling operations and the multiple trimming operations are applied to the image in an order in which one upscaling operation is applied to the image and thereafter one trimming operation corresponding to the upscaling operation is applied to the image.

10. The non-transitory computer-readable storage medium 8 , wherein the bitstream further comprising:

ratio information, wherein

the ratio information includes ratios of multiple downscaling operations performed to generate the bitstream or reciprocals of the ratios of the multiple downscaling operations performed to generate the bitstream.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 25, 2023
From: LEE, JOO-YOUNG; JEONG, SE-YOON; KWON, HYOUNG-JIN; KIM, DONG-HYUN; KIM, YOUN-HEE; KIM, JONG-HO; LEE, TAE-JIN; CHOI, JIN-SOO
To: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE
Reel/Frame 065013/0890 →
Priority Claims (2)
KR 10-2020-0027803 · Mar 5, 2020 · national
KR 10-2021-0026795 · Feb 26, 2021 · national
Continuity (2)
Continuation 17192480 · Mar 4, 2021
Related Publication 20240013449A1 · Jan 11, 2024
References Cited (11)
US 10321155B2 · Yoo et al. · 2019 [cited by applicant]
US 10841577B2 · Cho et al. · 2020 [cited by applicant]
US 20170104993A1 · Jeong et al. · 2017 [cited by applicant]
US 20180249158A1 · Huang et al. · 2018 [cited by applicant]
US 20190246102A1 · Cho et al. · 2019 [cited by applicant]
EP 3567857A1 · 2019 [cited by applicant]
KR 101663668B1 · 2016 [cited by applicant]
KR 1020190096281A · 2019 [cited by applicant]
Queiroz (“Processing JPEG-Compressed Images and documents”, IEEE 1998) (Year: 1998). [cited by examiner]
Anthony Geglio et al., Deep Convolutional Autoencoder for Assessment of Anomalies in Multi-stream Sensor Data, Feb. 15, 2022. [cited by applicant]
Olaf Ronneberger et al., U-Net: Convolutional Networks for Biomedical Image Segmentation, Springer 2015 (Year: 2015), pp. 235-241, Computer Science Department and BIOSS Centre for Biological Signalling Studies, Universi… [cited by applicant]