On Out-of-Distribution Detection Algorithms with Deep Neural Skin Cancer Classifiers

Pacheco, André G. C.; Sastry, Chandramouli Shama; Trappenberg, Thomas; Oore, Sageev; Krohling, Renato A.

doi:10.1109/cvprw50498.2020.00374

Cited by 33 publications

(27 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They found this pattern using different orders of gram matrices. Pachecho et al, [27] builds upon Gram OoD [31] and detected OoD samples for skin cancer. In their work, layerwise deviation of a sample's feature from its distribution is considered to be an indication of OoD and they have shown reasonable results on skin cancer-related OoD datasets.…”

Section: Related Workmentioning

confidence: 99%

“…We evaluate our approach using the setup proposed by [27], where two standard skin cancer datasets, ISIC 2018 and ISIC 2019 [11,35,12] are used as ID data. ISIC 2019 dataset contains 25,331 images of eight skin cancer classes i.e., Melanoma (MEL), Melanocytic nevus (NV), Basal cell carcinoma (BCC), Actinic keratosis (AK), Benign keratosis (BKL), Dermatofibroma (DF), Squamous cell carcinoma (SCC) and Vascular lesion Squamous cell carcinoma (VASC).…”

Section: Isic 2019 Datasetmentioning

confidence: 99%

“…We use six different OoD datasets [27] for the model trained on skin images of ISIC. The details of each of them are as follows:…”

Section: Isic 2019 Datasetmentioning

confidence: 99%

“…Experiments have shown that adding small noise to an image causes the network to incorrectly predict with high probability [17]. Similarly, as shown in [27], the image containing a dog can wrongly be classified as a skin cancer image using recent CNN architectures such as DenseNet, MobileNet, ResNet, and VGGNet, etc. This may result in catastrophic failures in real-world applications, especially in waqas.sultani@itu.edu.pk (W. Sultani) ORCID(s): healthcare systems.…”

Section: Introductionmentioning

confidence: 99%

“…Finally, instead of thresholding distance to the naive nearest neighbor, inspired by Kreciprocal nearest neighbors [40], we introduce a new robust OoD samples ranking strategy. We have compared the proposed approach with five state-of-the-art approaches on six OoD datasets [27] for skin cancer and five OoD datasets for malaria detection and have obtained promising results on all evaluation metrics. In summary, the proposed approach has the following contributions:…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Out of distribution detection for skin and malaria images

Zaida¹,

Ali²,

Ali³

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep neural networks have shown promising results in disease detection and classification using medical image data. However, they still suffer from the challenges of handling real-world scenarios especially reliably detecting out-of-distribution (OoD) samples. We propose an approach to robustly classify OoD samples in skin and malaria images without the need to access labeled OoD samples during training. Specifically, we use metric learning along with logistic regression to force the deep networks to learn much rich class representative features. To guide the learning process against the OoD examples, we generate ID similar-looking examples by either removing class-specific salient regions in the image or permuting image parts and distancing them away from in-distribution samples. During inference time, the K-reciprocal nearest neighbor is employed to detect out-of-distribution samples. For skin cancer OoD detection, we employ two standard benchmark skin cancer ISIC datasets as ID, and six different datasets with varying difficulty levels were taken as out of distribution. For malaria OoD detection, we use the BBBC041 malaria dataset as ID and five different challenging datasets as out of distribution. We achieved state-of-the-art results, improving 5% and 4% in TNR@ TPR95% over the previous state-of-the-art for skin cancer and malaria OoD detection respectively.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Isic 2019 Datasetmentioning

confidence: 99%

“…We use six different OoD datasets [27] for the model trained on skin images of ISIC. The details of each of them are as follows:…”

Section: Isic 2019 Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Out of distribution detection for skin and malaria images

Zaida¹,

Ali²,

Ali³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

Out-of-Distribution Detection for Long-Tailed and Fine-Grained Skin Lesion Images

Mehta

Gal²,

Bowling³

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

A Multi-scale Framework for Out-of-Distribution Detection in Dermoscopic Images

Huang

Wang

Cai

et al. 2023

Machine Learning for Cyber Security

View full text Add to dashboard Cite

The automatic detection of skin diseases via dermoscopic images can improve the efficiency in diagnosis and help doctors make more accurate judgments. However, conventional skin disease recognition systems may produce high confidence for outof-distribution (OOD) data, which may become a major security vulnerability in practical applications. In this paper, we propose a multi-scale detection framework to detect out-of-distribution skin disease image data to ensure the robustness of the system. Our framework extracts features from different layers of the neural network. In the early layers, rectified activation is used to make the output features closer to the well-behaved distribution, and then an one-class SVM is trained to detect OOD data; in the penultimate layer, an adapted Gram matrix is used to calculate the features after rectified activation, and finally the layer with the best performance is chosen to compute a normality score. Experiments show that the proposed framework achieves superior performance when compared with other state-of-the-art methods in the task of skin disease recognition.

show abstract

On Out-of-Distribution Detection Algorithms with Deep Neural Skin Cancer Classifiers

Cited by 33 publications

References 26 publications

Out of distribution detection for skin and malaria images

Out of distribution detection for skin and malaria images

Out-of-Distribution Detection for Long-Tailed and Fine-Grained Skin Lesion Images

A Multi-scale Framework for Out-of-Distribution Detection in Dermoscopic Images

Contact Info

Product

Resources

About