Samsung patents method to adjust image compression using neural networks
This patent describes a way to decode compressed images by using a neural network to determine how to unpack the data. It adjusts the statistical rules used to read the file based on specific details found in the image itself, rather than using a fixed set of rules.
In the filing’s own words
An image decoding method includes obtaining, second feature data from a bitstream for first feature data of a current image, obtaining quantization data and probability data by applying the second feature data to a neural network, modifying the probability data, based on sample values of the quantization data, obtaining quantized first feature data by applying entropy decoding, based on the modified probability data, on bits included in the bitstream, obtaining dequantized first feature data by dequantizing the quantized first feature data based on the sample values of the quantization data, and reconstructing the current image by performing neural network-based decoding on the dequantized first feature data. The first feature data having been obtained by performing neural network-based encoding on the current image.
What exactly they claim (claim 1)
1. An image decoding method, comprising: obtaining second feature data from a bitstream for first feature data, the first feature data having been obtained by performing neural network-based encoding on a current image; obtaining quantization data and probability data by applying the second feature data to a neural network; modifying the probability data, based on sample values of the quantization data; obtaining quantized first feature data by applying entropy decoding, based on the modified probability data, on bits comprised in the bitstream; obtaining dequantized first feature data by dequantizing the quantized first feature data based on the sample values of the quantization data; and reconstructing the current image by performing neural network-based decoding on the dequantized first feature data.
Facts
- Company
- Samsung (SAMSUNG ELECTRONICS CO., LTD.)
- Official title
- Image decoding device and image encoding device for adaptive quantization and inverse quantization, and method performed thereby
- Patent number
- US12739379B2
- Application number
- 19056172
- Filed
- Feb 18, 2025
- Granted
- Sep 15, 2026
- Inventors
- 2
- Claims
- 14
Classified under
H04N 19/124 Electricity & electronicsH04N 19/13 Electricity & electronicsH04N 19/137 Electricity & electronicsH04N 19/189 Electricity & electronicsMore from Samsung in AI & software
- Samsung patents tool to erase objects from photos using AI Sep 24, 2026
- Samsung patents drawing objects into photos with AI Oct 15, 2026
- Samsung patents AI that creates personalized exercise videos Sep 10, 2026
- Samsung patents AI summaries for paused TV shows based on your viewing habits Sep 24, 2026
- Samsung patents turning chat conversations into image prompts Oct 1, 2026
Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.