TRAMA
This 2-volume set LNCS 15297-15298 constitutes the refereed proceedings of the 46th Annual Conference of the German Association for Pattern Recognition, DAGM-GCPR 2024, held in Munich, Germany, during September 10-13, 2024.The 44 full papers included in these proceedings were carefully reviewed and selected from 81 submissions. They are organized in these topical sections:Part I: Clustering and Segmentation; Learning Techniques; Medical and Biological Applications; Uncertainty and Explainability.Part II: Modelling of Faces and Shapes; Image Generation and Reconstruction; 3D Analysis and Sythesis; Video Analysis; Photogrammetry and Remote Sensing.
SOMMARIO
.- Clustering and Segmentation. .- PARMESAN: Parameter-Free Memory Search and Transduction for Dense Prediction Tasks. .- A State-of-the-Art Cutting Plane Algorithm for Clique Partitioning. .- Self-Supervised Semantic Segmentation from Audio-Visual Data. .- BTSeg: Barlow Twins Regularization for Domain Adaptation in Semantic Segmentation. .- Learning Techniques. .- FullCert: Deterministic End-to-End Certification for Training and Inference of Neural Networks. .- Self-Masking Networks for Unsupervised Adaptation. .- A Theoretical Formulation on the Use of Multiple Positive Views in Contrastive Learning .- Decoupling of neural network calibration measures. .- Examining Common Paradigms in Multi-Task Learning. .- DIAGen: Semantically Diverse Image Augmentation with Generative Models for Few-Shot Learning. .- Efficient and Discriminative Image Feature Extraction for Universal Image Retrieval .. .- Anomaly Detection with Conditioned Denoising Diffusion Models. .- Medical and Biological Applications. .- SurgeoNet: Realtime 3D Pose Estimation of Articulated Surgical Instruments from Stereo Images using a Synthetically-trained Network. .- Foundation Models Permit Retinal Layer Segmentation Across OCT Devices. .- Correlation Clustering of Organoid Images. .- Animal Identification with Independent Foreground and Background Modeling. .- Robust Tumor Segmentation with Hyperspectral Imaging and Graph Neural Networks. .- Bigger Isn’t Always Better: Towards a General Prior for Medical Image Reconstruction.