Visual defect obfuscation based self-supervised anomaly detection.

Due to scarcity of anomaly situations in the early manufacturing stage, an unsupervised anomaly detection (UAD) approach is widely adopted which only uses normal samples for training. This approach is based on the assumption that the trained UAD model will accurately reconstruct normal patterns but struggles with unseen anomalies. To enhance the UAD performance, reconstruction-by-inpainting based methods have recently been investigated, especially on the masking strategy of suspected defective regions. However, there are still issues to overcome: (1) time-consuming inference due to multiple masking, (2) output inconsistency by random masking, and (3) inaccurate reconstruction of normal patterns for large masked areas. Motivated by this, this study proposes a novel reconstruction-by-inpainting method, dubbed Excision And Recovery (EAR), that features single deterministic masking based on the ImageNet pre-trained DINO-ViT and visual obfuscation for hint-providing. Experimental results on the MVTec AD dataset show that deterministic masking by pre-trained attention effectively cuts out suspected defective regions and resolves the aforementioned issues 1 and 2. Also, hint-providing by mosaicing proves to enhance the performance than emptying those regions by binary masking, thereby overcomes issue 3. The proposed approach achieves a high performance without any change of the model structure. Promising results are shown through laboratory tests with public industrial datasets. To suggest EAR be possibly adopted in various industries as a practically deployable solution, future steps include evaluating its applicability in relevant manufacturing environments.

Real-Time Deep Recognition of Standardized Liver Ultrasound Scan Locations.

Liver ultrasound (US) plays a critical role in diagnosing liver diseases. However, it is often difficult for examiners to accurately identify the liver segments captured in US images due to patient variability and the complexity of US images. Our study aim is automatic, real-time recognition of standardized US scans coordinated with reference liver segments to guide examiners. We propose a novel deep hierarchical architecture for classifying liver US images into 11 standardized US scans, which has yet to be properly established due to excessive variability and complexity. We address this problem based on a hierarchical classification of 11 US scans with different features applied to individual hierarchies as well as a novel feature space proximity analysis for handling ambiguous US images. Experiments were performed using US image datasets obtained from a hospital setting. To evaluate the performance under patient variability, we separated the training and testing datasets into distinct patient groups. The experimental results show that the proposed method achieved an F1-score of more than 93%, which is more than sufficient for a tool to guide examiners. The superior performance of the proposed hierarchical architecture was demonstrated by comparing its performance with that of non-hierarchical architecture.

Exploiting an Intermediate Latent Space between Photo and Sketch for Face Photo-Sketch Recognition.

The photo-sketch matching problem is challenging because the modality gap between a photo and a sketch is very large. This work features a novel approach to the use of an intermediate latent space between the two modalities that circumvents the problem of modality gap for face photo-sketch recognition. To set up a stable homogenous latent space between a photo and a sketch that is effective for matching, we utilize a bidirectional (photo → sketch and sketch → photo) collaborative synthesis network and equip the latent space with rich representation power. To provide rich representation power, we employ StyleGAN architectures, such as StyleGAN and StyleGAN2. The proposed latent space equipped with rich representation power enables us to conduct accurate matching because we can effectively align the distributions of the two modalities in the latent space. In addition, to resolve the problem of insufficient paired photo/sketch samples for training, we introduce a three-step training scheme. Extensive evaluation on a public composite face sketch database confirms superior performance of the proposed approach compared to existing state-of-the-art methods. The proposed methodology can be employed in matching other modality pairs.

Depth Edge Filtering Using Parameterized Structured Light Imaging.

This research features parameterized depth edge detection using structured light imaging that exploits a single color stripes pattern and an associated binary stripes pattern. By parameterized depth edge detection, we refer to the detection of all depth edges in a given range of distances with depth difference greater or equal to a specific value. While previous research has not properly dealt with shadow regions, which result in double edges, we effectively remove shadow regions using statistical learning through effective identification of color stripes in the structured light images. We also provide a much simpler control of involved parameters. We have compared the depth edge filtering performance of our method with that of the state-of-the-art method and depth edge detection from the Kinect depth map. Experimental results clearly show that our method finds the desired depth edges most correctly while the other methods cannot.

An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition.

This research features object recognition that exploits the context of object-action interaction to enhance the recognition performance. Since objects have specific usages, and human actions corresponding to these usages can be associated with these objects, human actions can provide effective information for object recognition. When objects from different categories have similar appearances, the human action associated with each object can be very effective in resolving ambiguities related to recognizing these objects. We propose an efficient method that integrates human interaction with objects into a form of object recognition. We represent human actions by concatenating poselet vectors computed from key frames and learn the probabilities of objects and actions using random forest and multi-class AdaBoost algorithms. Our experimental results show that poselet representation of human actions is quite effective in integrating human action information into object recognition.

Structural and Functional Brain Connectivity of People with Obesity and Prediction of Body Mass Index Using Connectivity.

Obesity is a medical condition affecting billions of people. Various neuroimaging methods including magnetic resonance imaging (MRI) have been used to obtain information about obesity. We adopted a multi-modal approach combining diffusion tensor imaging (DTI) and resting state functional MRI (rs-fMRI) to incorporate complementary information and thus better investigate the brains of non-healthy weight subjects. The objective of this study was to explore multi-modal neuroimaging and use it to predict a practical clinical score, body mass index (BMI). Connectivity analysis was applied to DTI and rs-fMRI. Significant regions and associated imaging features were identified based on group-wise differences between healthy weight and non-healthy weight subjects. Six DTI-driven connections and 10 rs-fMRI-driven connectivities were identified. DTI-driven connections better reflected group-wise differences than did rs-fMRI-driven connectivity. We predicted BMI values using multi-modal imaging features in a partial least-square regression framework (percent error 15.0%). Our study identified brain regions and imaging features that can adequately explain BMI. We identified potentially good imaging biomarker candidates for obesity-related diseases.

Masked fake face detection using radiance measurements.

This research presents a novel 2D feature space where real faces and masked fake faces can be effectively discriminated. We exploit the reflectance disparity based on albedo between real faces and fake materials. The feature vector used consists of radiance measurements of the forehead region under 850 and 685 nm illuminations. Facial skin and mask material show linearly separable distributions in the feature space proposed. By simply applying Fisher's linear discriminant, we have achieved 97.78% accuracy in fake face detection. Our method can be easily implemented in commercial face verification systems.

Effective representation using ICA for face recognition robust to local distortion and partial occlusion.

The performance of face recognition methods using subspace projection is directly related to the characteristics of their basis images, especially in the cases of local distortion or partial occlusion. In order for a subspace projection method to be robust to local distortion and partial occlusion, the basis images generated by the method should exhibit a part-based local representation. We propose an effective part-based local representation method named locally salient ICA (LS-ICA) method for face recognition that is robust to local distortion and partial occlusion. The LS-ICA method only employs locally salient information from important facial parts in order to maximize the benefit of applying the idea of "recognition by parts." It creates part-based local basis images by imposing additional localization constraint in the process of computing ICA architecture I basis images. We have contrasted the LS-ICA method with other part-based representations such as LNMF (Localized Nonnegative Matrix Factorization) and LFA (Local Feature Analysis). Experimental results show that the LS-ICA method performs better than PCA, ICA architecture I, ICA architecture II, LFA, and LNMF methods, especially in the cases of partial occlusions and local distortions.