Cartoon Explanations of Image Classifiers
نویسندگان
چکیده
We present CartoonX (Cartoon Explanation), a novel model-agnostic explanation method tailored towards image classifiers and based on the rate-distortion (RDE) framework. Natural images are roughly piece-wise smooth signals—also called cartoon-like images—and tend to be sparse in wavelet domain. is first exploit this by requiring its explanations domain, thus extracting relevant part of an instead pixel-sparse regions. demonstrate that can reveal valuable explanatory information, particularly for misclassifications. Moreover, we show achieves lower distortion with fewer coefficients than state-of-the-art methods.
منابع مشابه
Cartoon+Texture Image Decomposition
The algorithm first proposed in [3] stems from a theory proposed by Yves Meyer in [1]. The cartoon+texture algorithm decomposes any image f into the sum of a cartoon part, u , where only the image contrasted shapes appear, and a textural v part with the oscillating patterns. Such a decomposition f=u+v is analogous to the classical signal processing low pass-high pass filter decomposition. Howev...
متن کاملSelection of Image Classifiers 1 Selection of Image Classifiers
An approach to classifier combination based on the concept of “dynamic classifier selection” is presented in this letter. The results are reported and they show that the proposed approach allows to develop effective image classification systems.
متن کاملDirectional Filters for Cartoon + Texture Image Decomposition
We present in this article a detailed analysis and implementation of the cartoon+texture decomposition algorithm proposed in [A. Buades, J.L. Lisani, “Directional filters for color cartoon + texture image and video decomposition”, Journal of Mathematical Imaging and Vision, 2015]. This method follows the approach proposed by [A. Buades, T. Le, J-M. Morel, L. Vese, “Cartoon+Texture Image Decompo...
متن کاملTotal Variation Based Image Cartoon-Texture Decomposition‡
This paper studies algorithms for decomposing a real image into the sum of cartoon and texture based on total variation minimization and secondorder cone programming (SOCP). The cartoon is represented as a function of bounded variation while texture (and noise) is represented by elements in the space of oscillating functions, as proposed by Yves Meyer. Our approach gives more accurate results t...
متن کاملCartoon Extraction Based on Anisotropic Image Classification
We propose a new approach for the extraction of cartoons from 2D aerial images. Particularly in city areas, these images are mainly characterized by rectangular geometries of locally varying orientation. The presented method is based on a joint classification of the shape orientation and a rectangular structure preserving prior in the restoration of image shapes. Mathematically, an anisotropic ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Lecture Notes in Computer Science
سال: 2022
ISSN: ['1611-3349', '0302-9743']
DOI: https://doi.org/10.1007/978-3-031-19775-8_26