Unet Residual Resnet34

--- license: mit tags: - pytorch - image-segmentation - unet - medical-imaging - resnet - polyp-segmentation datasets: - RGarrido03/kvasir-seg-augmented metrics: - dice - iou model-index: - name: UNet Residual ResNet34 results: - task: type: image-segmentation name: Image Segmentation dataset: name: Kvasir-SEG (augmented) type: RGarrido03/kvasir-seg-augmented metrics: - type: dice value: 0.8991 name: Dice Score - type: iou value: 0.8167 name: IoU Score --- # UNet Residual ResNet34 A U-Net architecture for polyp segmentation from colonoscopy images, combining: - **ResNet34** pretrained encoder (ImageNet weights, early layers frozen during training) - **Residual blocks** with Squeeze-and-Excitation (SE) attention in the decoder - **ASPP** (Atrous Spatial Pyramid Pooling) bottleneck for multi-scale context Inspired by the [ResUNet++](https://arxiv.org/abs/1911.07067) architecture. ## Model description The encoder is a ResNet34 backbone that extracts multi-scale features at 5 resolutions (H, H/2, H/4, H/8, H/16) with channel counts [64, 64, 128, 256, 512]. Channel adapters project backbone features to the decoder's expected dimensions before skip connections. The bottleneck applies a `ResidualSEBlock` followed by ASPP (rates [1, 6, 12, 18]) and a projection convolution, giving the model a wide receptive field without losing spatial detail. The decoder uses four `ResidualSEBlock` stages with transposed convolutions for upsampling and concatenated skip connections from the encoder. A final 1×1 convolution produces the binary segmentation logit map. | Component | Detail | |------------------|--------| | Input size | 256×256×3 | | Output size | 256×256×1 (logits) | | Encoder | ResNet34 (pretrained, ImageNet) | | Bottleneck | ResidualSEBlock + ASPP | | Decoder blocks | ResidualSEBlock + ConvTranspose2d | | Parameters | ~28M | ## Training Trained on the [Kvasir-SEG](https://datasets.simula.no/kvasir-seg/) dataset with 5× random augmentation (geometric flips/rotations + brightness/contrast/saturation jitter), giving ~5 400 training samples. | Hyperparameter | Value | |-------------------------|-------| | Epochs | 8 | | Batch size | 128 | | Optimizer | AdamW | | Learning rate | 1e-3 (linear decay) | | Weight decay | 0.01 | | Loss | 0.5 × BCE + 0.5 × Dice | ### Training metrics | Epoch | Train loss | Val loss | Val Dice | Val IoU | |-------...