Prior works have demonstrated that implicit representations trained only for reconstruction tasks typically generate encodings that are not useful for semantic tasks. In this work, we propose a method that contextualises the encodings of implicit representations, enabling their use in downstream tasks (e.g. semantic segmentation), without requiring access to the original training data or encoding network. Using an implicit representation trained for a reconstruction task alone, our contextualising module takes an encoding trained for reconstruction only and reveals meaningful semantic information that is hidden in the encodings, without compromising the reconstruction performance. With our proposed module, it becomes possible to pre-train implicit representations on larger datasets, improving their reconstruction performance compared to training on only a smaller labelled dataset, whilst maintaining their segmentation performance on the labelled dataset. Importantly, our method allows for future foundation implicit representation models to be fine-tuned on unseen tasks, regardless of encoder or dataset availability.

翻译：先前研究表明，仅通过重建任务训练的隐式表征通常生成的编码难以用于语义任务。本研究提出一种方法，对隐式表征的编码进行上下文化处理，使其可应用于下游任务（例如语义分割），且无需访问原始训练数据或编码网络。基于仅通过重建任务训练的隐式表征，我们的上下文化模块可提取编码中隐藏的语义信息，同时不降低重建性能。借助该模块，可在更大数据集上预训练隐式表征，相比仅在较小标注数据集上训练，其重建性能得到提升，同时在标注数据集上仍保持分割性能。重要的是，本方法允许未来的基础隐式表征模型针对未见任务进行微调，而无需考虑编码器或数据集的可用性。