20.500.12556/RUL-148675 Human-centered deep compositional model for handling occlusions Despite their powerful discriminative abilities, Convolutional Neural Networks (CNNs) lack the properties of generative models. This leads to a decreased performance in environments where objects are poorly visible. Solving such a problem by adding more training samples can quickly lead to a combinatorial explosion, therefore the underlying architecture has to be changed instead. This work proposes a Human-Centered Deep Compositional model (HCDC) that combines low-level visual discrimination of a CNN and the high-level reasoning of a Hierarchical Compositional model (HCM). Defined as a transparent model, it can be optimized to real-world environments by adding compactly encoded domain knowledge from human studies and physical laws. The new FridgeNetv2 dataset and a mixture of publicly available datasets are used as a benchmark. The experimental results show the proposed model is explainable, has higher discriminative and generative power, and better handles the occlusion than the current state-of-the-art Mask-RCNN in instance segmentation tasks. computer vision deep learning convolutional neural networks hierarchical compositional model occlusion discriminability generalizability interpretability domain knowledge instance segmentation occlusion handling računalniški vid globoko učenje konvolucijske nevronske mreže hierarhični kompozicionalni model zakrivanje interpretabilnost poznavanje področja true false true Angleški jezik Slovenski jezik Članek v reviji 2023-08-29 08:41:06 2023-08-29 08:41:07 2023-08-30 03:40:29 0000-00-00 00:00:00 2023 0 0 14 str. Vol. 138, art. 109397 2023 0000-00-00 Zaloznikova Objavljeno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004 0031-3203 10.1016/j.patcog.2023.109397 142438403 1-s2.0-S0031320323000985-main.pdf 1-s2.0-S0031320323000985-main.pdf 1 20AE65390FD3DB201D536E1AC518895F 2da68b8c4c4c3cb0dc04852f31fdd7338e748071168d47811912591233fb7ba4 74b36e50-4637-11ee-a605-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=172969 https://www.sciencedirect.com/science/article/pii/S0031320323000985 0 0747bad7-4637-11ee-a605-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=172968 Fakulteta za elektrotehniko 0 0 0