Deep Residual Learning for Image Recognition
Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun
TLDR
This paper introduces deep residual networks that enable training of substantially deeper neural networks by learning residual functions, leading to state-of-the-art performance in image recognition tasks.
Key contributions
- Proposes residual learning framework that reformulates layers to learn residual functions instead of direct mappings.
- Demonstrates successful training of very deep networks (up to 152 layers) with improved accuracy and lower complexity than previous models.
- Achieves top performance on ImageNet and COCO benchmarks, winning multiple 2015 ILSVRC and COCO competitions.
Why it matters
This paper matters because it overcomes the fundamental difficulty of training very deep neural networks by introducing residual connections, which significantly improve optimization and accuracy. This breakthrough enables the development of much deeper architectures that have become the foundation for many modern computer vision systems, pushing the state-of-the-art in image classification, detection, and segmentation.
Original Abstract
Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.