Publications by authors named "O Deforges"
Predictive Uncertainty Estimation for Camouflaged Object Detection.
IEEE Trans Image Process
July 2023
Article Synopsis
- Machine learning methods for camouflaged object detection face two main biases: model bias, caused by training datasets focusing on the image center, and data bias, due to challenges in accurately labeling object boundaries.
- To address these biases, a new technique called predictive uncertainty estimation is introduced, which combines model uncertainty and data uncertainty.
- The proposed Predictive Uncertainty Estimation Network (PUENet) effectively estimates these uncertainties using a Bayesian framework, achieving high prediction accuracy and reliable uncertainty estimation in experiments.
In this paper, we propose, implement, and analyze the structures of two keyed hash functions using the Chaotic Neural Network (CNN). These structures are based on Sponge construction, and they produce two variants of hash value lengths, i.e.
View Article and Find Full Text PDFIn this paper, we firstly study the security enhancement of three steganographic methods by using a proposed chaotic system. The first method, namely the Enhanced Edge Adaptive Image Steganography Based on LSB Matching Revisited (EEALSBMR), is present in the spatial domain. The two other methods, the Enhanced Discrete Cosine Transform (EDCT) and Enhanced Discrete Wavelet transform (EDWT), are present in the frequency domain.
View Article and Find Full Text PDFUltrasonographic examination, either as visual inspection or quantitative analysis, is the most widely diagnostic resource. However, speckle noise is one of the drawbacks that makes it less effective than other medical imaging systems. Several speckle reduction methods often offer effective speckle reduction but generally suffer from oversmoothing, a blurring effect and a man-made appearance.
View Article and Find Full Text PDFNIQSV+: A No-Reference Synthesized View Quality Assessment Metric.
IEEE Trans Image Process
April 2018
Article Synopsis
- The paper discusses challenges in assessing the quality of synthesized 3-D views due to unique distortions that arise from view synthesis and depth map compression, which traditional quality metrics can't address since original reference views are often unavailable.
- A new no-reference image quality assessment method called NIQSV+ is introduced, which evaluates synthesized views by identifying distortions like blurriness, black holes, and stretching without needing a reference image or depth map.
- Experimental results indicate that NIQSV+ performs comparably to the best full-reference metrics and significantly outperforms other no-reference metrics, effectively aligning with human subjective quality assessments.