A5052562698- Neural dynamics and brain function
- Visual Attention and Saliency Detection
- Industrial Vision Systems and Defect Detection
- Visual perception and processing mechanisms
- Image Processing Techniques and Applications
- Neural and Behavioral Psychology Studies
- Cell Image Analysis Techniques
- Integrated Circuits and Semiconductor Failure Analysis
- Neurotransmitter Receptor Influence on Behavior
- Memory and Neural Mechanisms
- Advanced Surface Polishing Techniques
- Retinal Imaging and Analysis
- Receptor Mechanisms and Signaling
- AI in cancer detection
- Infrared Target Detection Methodologies
- Advanced Memory and Neural Computing
- Image and Object Detection Techniques
- Brain Tumor Detection and Classification
- Glaucoma and retinal disorders
- Robot Manipulation and Learning
- Evolutionary Algorithms and Applications
- Advanced Neural Network Applications
- Face Recognition and Perception
- Retinal and Optic Conditions
- Retinal Diseases and Treatments
Chemnitz University of Technology
2012-2024
Abstract In the semiconductor industry, automated visual inspection aims to improve detection and recognition of manufacturing defects by leveraging power artificial intelligence computer vision systems, enabling manufacturers profit from an increased yield reduced costs. Previous domain-specific contributions often utilized classical approaches, whereas more novel systems deploy deep learning based ones. However, a persistent problem in domain stems very small defect patterns which are size...
Abstract In ophthalmology, intravitreal operative medication therapy (IVOM) is a widespread treatment for diseases related to the age-related macular degeneration (AMD), diabetic edema, as well retinal vein occlusion. However, in real-world settings, patients often suffer from loss of vision on time scales years despite therapy, whereas prediction visual acuity (VA) and earliest possible detection deterioration under real-life conditions challenging due heterogeneous incomplete data. this...
Can we attend to multiple distinct spatial locations at the same time? According a recent psychophysical study [J. Dubois et al. (2009)Journal of Vision, 9, 3.1-11] such split attention might be limited short periods time. Following N. P. Bichot [(1999)Perception & Psychophysics, 61, 403-423] subjects had report identity letters that were briefly presented different locations, while two these (targets) relevant for concurrent shape comparison task. In addition design used by stimulus onset...
Reaching a target object in an unknown and unstructured environment is easily performed by human beings. However, designing humanoid robot that executes the same task requires implementation of complex abilities, such as identifying visual field, estimating its spatial location, precisely driving motors arm to reach it. While research usually tackles development abilities singularly, this work we integrate number computational models into unified framework, demonstrate torso feasibility...
Modern parallel hardware such as multi-core processors (CPUs) and graphics processing units (GPUs) have a high computational power which can be greatly beneficial to the simulation of large-scale neural networks. Over past years, number efforts focused on developing algorithms simulators best suited for spiking models. In this article, we aim at investigating advantages drawbacks CPU GPU parallelization mean-firing rate neurons, widely used in systems-level neuroscience. By comparing OpenMP,...
It is a long-term goal to transfer biological processing principles as well the power of human recognition into machine vision and engineering systems. One such visual attention, smart concept which focuses on part scene. In this contribution, we utilize attention improve automatic detection defect patterns for wafers within domain semiconductor manufacturing. Previous works in have often utilized classical learning approaches KNNs, SVMs, or MLPs, while few already used modern like deep...
Automated visual inspection in the semiconductor industry aims to detect and classify manufacturing defects utilizing modern image processing techniques. While an earliest possible detection of defect patterns allows quality control automation chains, manufacturers benefit from increased yield reduced costs. Since classical systems are limited their ability novel patterns, machine learning approaches often involve a tremendous amount computational effort, this contribution introduces deep...
Whereas conventional state-of-the-art image processing systems of recording and output devices almost exclusively utilize square arranged methods, biological models, however, suggest an alternative, evolutionarily-based structure. Inspired by the human visual perception system, hexagonal in context machine learning offers a number key advantages that can benefit both researchers users alike. The deep framework Hexnet leveraged this contribution serves therefore generation images utilizing...
In semiconductor manufacturing, the wafer dicing process is central yet vulnerable to defects that significantly impair yield - proportion of defect-free chips. Deep neural networks are current state art in (semi-)automated visual inspection. However, they notoriously known require a particularly large amount data for model training. To address these challenges, we explore application generative adversarial (GAN) image augmentation and classification induced enhance variety balance training...
Computational models of visual attention have replicated a large number data from experiments. However, typically each computational model has been shown to account for only few sets. Thus, general fully understand the attentive dynamics in cortex is still missing. To reveal set principles that determine attentional selection cortex, we developed novel attention, particularly focused on explaining single cell recordings multiple brain areas. Among those are spatial- and feature-based biased...
In ophthalmology, intravitreal operative medication therapy (IVOM) is a widespread treatment for diseases related to the age-related macular degeneration (AMD), diabetic edema (DME), as well retinal vein occlusion (RVO). However, in real-world settings, patients often suffer from loss of vision on time scales years despite therapy, whereas prediction visual acuity (VA) and earliest possible detection deterioration under real-life conditions challenging due heterogeneous incomplete data. this...