用端到端训练提升SIREN在图像分类中的表现，超越现有方法。
End-to-End Implicit Neural Representations for Classification
- 通过元学习初始化SIREN并自适应调整学习率，直接优化参数表示
- 在CIFAR-10上无增强时准确率从38.8%提升至59.6%，有增强达64.7%
- 首次实现对ImageNet-1K的SIREN分类，准确率达23.6%，性能领先
隐式神经表示（INRs）如NeRF和SIREN通过网络参数编码信号，在信号重建任务中表现优异。但将其用于下游任务如分类仍面临挑战，因参数固有的对称性导致性能远低于基于像素的CNN方法。本文提出一种端到端策略，联合初始化SIREN与可学习的学习率方案，生成更优分类表示。我们发现，仅用一个简单的Transformer模型作用于元学习得到的SIREN，无需显式引入对称等变性设计，即可超越当前最先进水平。在无数据增强的CIFAR-10 SIREN分类任务中，准确率从38.8%提升至59.6%；在含增强条件下从63.4%提升至64.7%。在高分辨率Imagenette数据集上也表现出良好可扩展性，重建质量合理且分类准确率达60.8%。首次实现对完整ImageNet-1K的SIREN分类，达到23.6%的准确率。据我们所知，此前尚无其他SIREN分类方法在高分辨率数据集上建立基准。代码已开源。
原文摘要 · Abstract (English)
Implicit neural representations (INRs) such as NeRF and SIREN encode a signal in neural network parameters and show excellent results for signal reconstruction. Using INRs for downstream tasks, such as classification, is however not straightforward. Inherent symmetries in the parameters pose challenges and current works primarily focus on designing architectures that are equivariant to these symmetries. However, INR-based classification still significantly under-performs compared to pixel-based methods like CNNs. This work presents an end-to-end strategy for initializing SIRENs together with a learned learning-rate scheme, to yield representations that improve classification accuracy. We show that a simple, straightforward, Transformer model applied to a meta-learned SIREN, without incorporating explicit symmetry equivariances, outperforms the current state-of-the-art. On the CIFAR-10 SIREN classification task, we improve the state-of-the-art without augmentations from 38.8% to 59.6%, and from 63.4% to 64.7% with augmentations. We demonstrate scalability on the high-resolution Imagenette dataset achieving reasonable reconstruction quality with a classification accuracy of 60.8% and are the first to do INR classification on the full ImageNet-1K dataset where we achieve a SIREN classification performance of 23.6%. To the best of our knowledge, no other SIREN classification approach has managed to set a classification baseline for any high-resolution image dataset. Our code is available at https://github.com/SanderGielisse/MWT
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。