OpenCV reposted this
We're excited to announce the release of Anomalib v2.1.0! This version brings several state-of-the-art models and anomaly detection datasets. Check it out! 👇 🎉 New models : 🖼️ UniNet (CVPR 2025): A contrastive learning-guided unified framework with feature selection for anomaly detection. 🖼️ Dinomaly (CVPR 2025): A 'less is more philosophy' encoder-decoder architecture model leveraging pre-trained foundational models. 🎥 Fuvas (ICASSP 2025): Few-shot unsupervised video anomaly segmentation via low-rank factorization of spatio-temporal features. 🎉 New datasets: ✨ MVTec AD 2 : A new version of the MVTec AD dataset with 8 categories of industrial anomaly detection ✨ MVTec LOCO AD : MVTec logical constraints anomaly detection dataset that includes both structural and logical anomalies. ✨ Real-IAD : A real-world multi-view dataset for benchmarking versatile industrial anomaly detection. ✨ VAD dataset : Valeo Anomaly Dataset (VAD) showcasing a diverse range of defects, from highly obvious to extremely subtle. Check out the full Changelog: https://lnkd.in/ezfKxbXi Samet Akcay Rajesh Gangireddy Radwan Ibrahim Blaž Rolih Aimira Baitieva Alexander Barabanov Eugene Liu Alfie Roddan Open.Intel #anomalydetection #machinelearning #opensource #anomalib #intel