000			02135naa a2200289 a 4500
003			AR-LpUFIB
005			20250311170534.0
008			230201s2024 xx o 000 0 spa d
024	8		_aDIF-M8955 _b9183 _zDIF008217
040			_aAR-LpUFIB _bspa _cAR-LpUFIB
100	1		_aStanchi, Oscar
245	1	0	_aQuantitative evaluation of white & black box interpretability methods for image classification
300			_a1 archivo (1,37 MB)
500			_aFormato de archivo PDF. -- Este documento es producción intelectual de la Facultad de Informática - UNLP (Colección BIPA/Biblioteca)
520			_aThe field of interpretability in Deep Learning faces significant challenges due to the lack of standard metrics for systematically evaluating and comparing interpretability methods. The absence of quantifiable measures impedes practitioners ability to select the most suitable methods and models for their specific tasks. To address this issue, we propose the Pixel Erosion and Dilation Score, a novel metric designed to assess the robustness of model explanations. Our approach involves applying iterative erosion and dilation processes to heatmaps generated by various interpretability methods, thereby using them to hide and show the important regions of a image to the network, allowing for a coherent and interpretable evaluation of model decision-making processes. We conduct quantitative ablation tests using our metric on the ImageNet dataset with both VGG16 and ResNet18 models. The results reveal that our new measure provides a numerical and intuitive means for comparing interpretability methods and models, facilitating more informed decision-making for practitioner.
534			_aCongreso Argentino de Ciencias de la Computación (30mo : 2024 : La Plata, Argentina)
650		4	_aVISIÓN POR COMPUTADORA
653			_aaprendizaje profundo
700	1		_aRonchetti, Franco
700	1		_aDal Bianco, Pedro A.
700	1		_aRíos, Gastón Gustavo
700	1		_aHasperué, Waldo
700	1		_aPuig, Domenec
700	1		_aRashwan, Hatem
856	4	0	_uhttp://sedici.unlp.edu.ar/handle/10915/176288
942			_cCP
999			_c57986 _d57986