SLSDeep: Skin Lesion Segmentation Based on Dilated Residual and Pyramid Pooling Networks
This work addresses automated melanoma diagnosis by improving segmentation accuracy and speed, though it appears incremental as it builds on existing deep learning architectures.
The paper tackled skin lesion segmentation in dermoscopic images by proposing SLSDeep, an encoder-decoder network using dilated residual layers and a pyramid pooling network, which outperformed state-of-the-art methods on ISBI 2016 and 2017 datasets and processed over 100 images per second on a GPU.
Skin lesion segmentation (SLS) in dermoscopic images is a crucial task for automated diagnosis of melanoma. In this paper, we present a robust deep learning SLS model, so-called SLSDeep, which is represented as an encoder-decoder network. The encoder network is constructed by dilated residual layers, in turn, a pyramid pooling network followed by three convolution layers is used for the decoder. Unlike the traditional methods employing a cross-entropy loss, we investigated a loss function by combining both Negative Log Likelihood (NLL) and End Point Error (EPE) to accurately segment the melanoma regions with sharp boundaries. The robustness of the proposed model was evaluated on two public databases: ISBI 2016 and 2017 for skin lesion analysis towards melanoma detection challenge. The proposed model outperforms the state-of-the-art methods in terms of segmentation accuracy. Moreover, it is capable to segment more than $100$ images of size 384x384 per second on a recent GPU.