An image encoding apparatus of the present invention comprises a frequency converting unit which executes frequency conversion on the data of a frame in moving image data, and generates coefficient data of a plurality of subbands, an encoding unit which quantizes the coefficient data acquired by the frequency converting unit in accordance with a quantization parameter, and encodes the quantized coefficient data, and a code amount controlling unit which controls the encoding unit such that a code amount generated by the encoding unit achieves a target code amount are provided. Here, the code amount controlling unit determines a subband target code amount of each of the plurality of subbands by distributing the target code amount to the plurality of subbands based on a ratio determined according to an indicator indicating a difficulty calculated for each of the plurality of subbands.

A transcoding service is described that is capable of transcoding or otherwise processing content, such as video, audio or multimedia content, by utilizing one or more pipelines. A pipeline can enable a user to submit transcoding jobs (or other processing jobs) into an available pipeline, where a transcoding service (or other such service) assigns one or more computing resources to process the jobs received to each pipeline. The transcoding service and the pipelines can be provided by at least one service provider (e.g., a cloud computing provider) or other such entity to a plurality of customers. A service provider can also provide the computing resources (e.g., servers, virtual machines, etc.) used to process the transcoding jobs from the pipelines.

This disclosure relates generally to distributed video coding. In one embodiment, distributed video encoding apparatus to achieve improved rate distortion performance is disclosed. The distributed video encoding apparatus comprises a processor and a memory communicatively coupled to the processor. The memory stores processor instructions, which, on execution, causes the processor to receive at least one Group of Pictures (GOP) comprising at least one key frame and at least one Wyner-Ziv (WZ) frame. The processor further determines a first value that is indicative of a cumulative motion activity associated with the at least one GOP. The processor further classifies the at least one GOP into one of one or more high-motion WZ frames and one or more low-motion WZ frames based on the determined first value. The processor encodes the high-motion WZ frames using inter no-motion encoding. The processor further encodes the one or more low-motion WZ frames using Wyner-Ziv encoding.

A video image decoding apparatus includes a plurality of predicted image generating units that generate predicted images according to respectively different methods. A predicted image combining unit combines the predicted images generated by the predicted image generating units to obtain a predicted image. A decoding unit decodes an encoded image by using, as side information, the predicted image obtained by the predicted image combining unit.

A method comprising: obtaining a block of a picture or a picture in an encoder; determining if the block/picture is used for on-line learning; if affirmative, encoding the block/picture; reconstructing a coarse version of the block/picture or the respective prediction error block/picture; enhancing the coarse version using a neural net; fine-tuning the neural net with a training signal based on the coarse version; determining if the block/picture is enhanced using the neural net; and if affirmative, encoding the block/picture with enhancing using the neural net.

Techniques for managing the assignment of transcoding tasks to transcoding nodes in a transcoding system are described. In one embodiment, for example, an apparatus may comprise circuitry and a transcoding management module for execution on the circuitry to assign a transcoding task to one of a set of transcoding nodes based on a set of task characteristics of the transcoding task and a set of efficiency values for the set of transcoding nodes, each of the set of efficiency values corresponding to a respective one of the set of transcoding nodes. Other embodiments are described and claimed.

Aspects of the disclosure provide methods and apparatuses for neural network processing, such as in video processing. In some examples, an apparatus for neural network processing includes processing circuitry. The processing circuitry determines that an input for a convolution operation includes a first input channel that is piecewise constant. Then, the processing circuitry calculates a first intermediate output channel based on other channels of the input for the convolution operation; and then generates an output of the convolution operation based on a combination (e.g., a linear combination) of the first intermediate output channel and the first input channel.

Aspects of the disclosure provide methods and apparatuses for neural network processing, such as in video processing. In some examples, an apparatus for neural network processing includes processing circuitry. The processing circuitry determines that an input for a convolution operation includes a first input channel that is piecewise constant. Then, the processing circuitry calculates a first intermediate output channel based on other channels of the input for the convolution operation; and then generates an output of the convolution operation based on a combination (e.g., a linear combination) of the first intermediate output channel and the first input channel.

A method comprising: obtaining a block of a picture or a picture in an encoder; determining if the block/picture is used for on-line learning; if affirmative, encoding the block/picture; reconstructing a coarse version of the block/picture or the respective prediction error block/picture; enhancing the coarse version using a neural net; fine-tuning the neural net with a training signal based on the coarse version; determining if the block/picture is enhanced using the neural net; and if affirmative, encoding the block/picture with enhancing using the neural net.

Embodiments of this application disclose a picture transmission method performed at a computer device. After obtaining a picture, the computer device generates a video sequence by replicating the picture N times, and N being a positive integer. Next the computer device obtains a resolution setting sequence and encodes the N to-be-encoded pictures in the video sequence according to the resolution setting sequence to generate N encoded pictures, each encoded picture having an associated resolution setting. Finally, the computer device sends the N encoded pictures to a decoding computer device. The decoding computer device then decodes and displays the N encoded pictures according to their respective resolution settings from low to high.