信息技术 多媒体内容描述接口 第17部分：用于多媒体内容描述和分析的神经网络压缩

[中文适用范围]： 本文件规定了神经网络编码 (NNC) 作为经过训练的神经网络的参数/权重的压缩表示，以及压缩表示的解码过程，补充了现有神经网络 (交换) 格式中的网络拓扑描述。它建立了一个压缩方法工具箱，指定 (在适用的情况下) 压缩比特流的结果元素。这些工具中的大多数可以应用于整个神经网络的压缩，其中一些还可以应用于压缩神经网络相对于基础网络的差异更新。例如，当模型在微调或迁移学习后重新分布时，或者当提供具有不同压缩比的神经网络版本时，这种差异更新很有用。本文件没有规定神经网络传输的完整协议，而是侧重于网络参数的压缩。仅指定语法格式、语义、相关解码过程要求、参数稀疏化、参数变换方法、参数量化、熵编码方法和现有交换格式内的集成/信令，而预处理、系统信令和复用、数据丢失恢复和后处理等其他事项被认为超出了本文档的范围。此外，解码器内执行的内部处理步骤也被认为超出了本文档的范围；只有外部可观察的输出行为才需要符合本文档的规范。 [外文原描述]： This document specifies Neural Network Coding (NNC) as a compressed representation of the parameters/weights of a trained neural network and a decoding process for the compressed representation, complementing the description of the network topology in existing (exchange) formats for neural networks. It establishes a toolbox of compression methods, specifying (where applicable) the resulting elements of the compressed bitstream. Most of these tools can be applied to the compression of entire neural networks, and some of them can also be applied to the compression of differential updates of neural networks with respect to a base network. Such differential updates are for example useful when models are redistributed after fine-tuning or transfer learning, or when providing versions of a neural network with different compression ratios. This document does not specify a complete protocol for the transmission of neural networks, but focuses on compression of network parameters. Only the syntax format, semantics, associated decoding process requirements, parameter sparsification, parameter transformation methods, parameter quantization, entropy coding method and integration/signalling within existing exchange formats are specified, while other matters such as pre-processing, system signalling and multiplexing, data loss recovery and post-processing are considered to be outside the scope of this document. Additionally, the internal processing steps performed within a decoder are also considered to be outside the scope of this document; only the externally observable output behaviour is required to conform to the specifications of this document.