FastFlow adalah metode deteksi dan lokalisasi anomali tanpa pengawasan yang menggunakan 2D normalizing flows sebagai estimator distribusi probabilitas, mampu dipasangkan dengan berbagai ekstraktor fitur seperti ResNet dan Vision Transformer untuk mencapai akurasi tinggi pada dataset MVTec AD.
FastFlow mengimplementasikan 2D normalizing flows sebagai estimator distribusi probabilitas yang dipasangkan dengan ekstraktor fitur deep learning (ResNet atau Vision Transformer). Selama pelatihan, model belajar mentransformasi fitur visual input ke distribusi normal standar, sehingga fitur area normal berada di pusat distribusi sedangkan fitur anomali berada jauh dari pusat. Evaluasi dilakukan secara ekstensif pada dataset benchmark MVTec AD menggunakan berbagai konfigurasi backbone.
Dibuat 19 Mei 2026 · paper_summarizer_v1.0
Buka PDF di perangkat Anda
Unduh / Buka PDFUnsupervised anomaly detection and localization is crucial to the practical application when collecting and labeling sufficient anomaly data is infeasible. Most existing representation-based approaches extract normal image features with a deep convolutional neural network and characterize the corresponding distribution through non-parametric distribution estimation methods. The anomaly score is calculated by measuring the distance between the feature of the test image and the estimated distribution. However, current methods can not effectively map image features to a tractable base distribution and ignore the relationship between local and global features which are important to identify anomalies. To this end, we propose FastFlow implemented with 2D normalizing flows and use it as the probability distribution estimator. Our FastFlow can be used as a plug-in module with arbitrary deep feature extractors such as ResNet and vision transformer for unsupervised anomaly detection and localization. In training phase, FastFlow learns to transform the input visual feature into a tractable distribution and obtains the likelihood to recognize anomalies in inference phase. Extensive experimental results on the MVTec AD dataset show that FastFlow surpasses previous state-of-the-art methods in terms of accuracy and inference efficiency with various backbone networks. Our approach achieves 99.4% AUC in anomaly detection with high inference efficiency.
