Methods and equipment for encoding video and computer-readable storage media that is not temporary

Abstract

The invention relates to a method and apparatus for video encoding. This apparatus comprises a processing circuit structured to generate an initial characteristic representation from the input image to be encoded and to perform a repetitive update of the values ​​of the elements in the initial characteristic representation. Iterative updates involve generating an encoded representation that corresponds to the final feature representation based on the final feature representation that has been updated from the initial feature representation by the number of iterations of the iterative update*. The reconstructed image that corresponds to the final feature representation is generated based on the encoded representation. The encoded image that corresponds to the final feature representation has updated values ​​of the generated elements. One of the following is (i) the scaled distortion loss that corresponds to the final characteristic representation or (ii) the number of repetitions of the repeated update that satisfies a predefined condition.