Publications by authors named "Bekkouch Imad Eddine Ibrahim"
Morphological abnormalities of the femoroacetabular (hip) joint are among the most common human musculoskeletal disorders and often develop asymptomatically at early easily treatable stages. In this paper, we propose an automated framework for landmark-based detection and quantification of hip abnormalities from magnetic resonance (MR) images. The framework relies on a novel idea of multi-landmark environment analysis with reinforcement learning.
View Article and Find Full Text PDFDetecting objects with a small representation in images is a challenging task, especially when the style of the images is very different from recent photos, which is the case for cultural heritage datasets. This problem is commonly known as few-shot object detection and is still a new field of research. This article presents a simple and effective method for black box few-shot object detection that works with all the current state-of-the-art object detection models.
View Article and Find Full Text PDFAnomaly Detection Based on Zero-Shot Outlier Synthesis and Hierarchical Feature Distillation.
IEEE Trans Neural Netw Learn Syst
January 2022
Article Synopsis
- Anomaly detection faces challenges due to the rarity of anomalies, which leads to unbalanced data issues; synthetic anomalies are proposed as a potential solution for this problem.
- The article introduces a two-level hierarchical latent space representation using autoencoders to create robust feature representations for generating synthetic anomalies without prior examples.
- The proposed method successfully generates pseudo outlier samples, enabling the training of effective binary classifiers for real anomaly detection, and shows strong performance across multiple benchmarking tests.