Deep Residual Networks (ResNet) ImageNet Classification
Classifies images into 1,000 ImageNet categories using residual learning with skip connections. Three variants available: ResNet-18 (69.57% top-1), ResNet-50 (75.99% top-1), ResNet-101 (77.56% top-1). Trade-off accuracy for speed/parameters.
What it does
ResNet classifies images into 1,000 ImageNet categories using residual learning (skip connections). Architecture enables training of very deep networks (18–101 layers) without vanishing gradient problems. Three model sizes available for different accuracy/efficiency requirements.
Problem it solves
- Enable training of very deep networks without gradient degradation
- Provide scalable accuracy-efficiency trade-off (lightweight to high-accuracy variants)
- Foundation for downstream tasks (object detection, localization, segmentation)
Input/Output
Accuracy & Speed
Model | Top-1 Accuracy | Top-5 Accuracy | Parameters |
ResNet-18 | 69.57% | 89.24% | 11.7M |
ResNet-50 | 75.99% | 92.98% | 25.6M |
ResNet-101 | 77.56% | 93.79% | 44.5M |
Best For:
- Lightweight: ResNet-18 (speed-focused, mobile/edge devices)
- Balanced: ResNet-50 (standard, good accuracy-speed trade-off)
- High-Accuracy: ResNet-101 (accuracy-focused, server/research)
Technical Details
Architecture | Residual Learning with Skip Connections |
Core Innovation: y = F(x) + x | • F(x): Residual function learned by stacked layers
• x: Skip connection (identity bypass)
• Benefits: Enables gradient flow through deep networks, mitigates vanishing gradient |
Building block | • ResNet-18: Basic blocks (2 conv layers per block)
• ResNet-50/101: Bottleneck blocks (1×1 reduce → 3×3 → 1×1 restore) |
Training | • Dataset: ImageNet-1k (1.28M training images, 1,000 classes)
• Optimizer: SGD (momentum 0.9)
• Batch size: 256
• Learning rate: 0.1 (÷10 every 30 epochs)
• Epochs: 100
• Weight decay: 1e-4
• Data augmentation:
◦ Random crop 224×224
◦ Random horizontal flip
◦ Multi-scale jitter (256–480)
◦ ImageNet normalization (per-channel mean/std) |
Depth Per Stage | • ResNet-18: [2, 2, 2, 2]
• ResNet-50: [3, 4, 6, 3]
• ResNet-101: [3, 4, 23, 3] |
Compliance & Provenance
Provider | Open-source |
Provider type | Specialized |
License | Apache 2.0 |
EU AI Act risk class | Minimal Risk |
Art. 50 transparency | Not applicable |
Region availability | Available globally |
Training data summary | Pending — provider has not yet published per Art. 53(d) |
For more on how we classify models and mark outputs, see our AI Policy.
Model Source
- HuggingFace:
- ResNet-18: https://huggingface.co/microsoft/resnet-18
- ResNet-50: https://huggingface.co/microsoft/resnet-50
- ResNet-101: https://huggingface.co/microsoft/resnet-101
- Official Implementation: https://github.com/KaimingHe/deep-residual-networks
- Paper: He et al., "Deep Residual Learning for Image Recognition" CVPR 2016 (Best Paper Award) https://arxiv.org/pdf/1512.03385
- License: apache-2.0