The goals of this research were to search for Convolutional Neural Network (CNN) architectures, suitable for an on-device processor with limited computing resources, performing at substantially lower Network Architecture Search (NAS) costs. A new algorithm entitled an Early Exit Population Initialisation (EE-PI) for Evolutionary Algorithm (EA) was developed to achieve both goals. The EE-PI reduces the total number of parameters in the search process by filtering the models with fewer parameters than the maximum threshold. It will look for a new model to replace those models with parameters more than the threshold. Thereby, reducing the number of parameters, memory usage for model storage and processing time while maintaining the same performance or accuracy. The search time was reduced to 0.52 GPU day. This is a huge and significant achievement compared to the NAS of 4 GPU days achieved using NSGA-Net, 3,150 GPU days by the AmoebaNet model, and the 2,000 GPU days by the NASNet model. As well, Early Exit Evolutionary Algorithm networks (EEEA-Nets) yield network architectures with minimal error and computational cost suitable for a given dataset as a class of network algorithms. Using EEEA-Net on CIFAR-10, CIFAR-100, and ImageNet datasets, our experiments showed that EEEA-Net achieved the lowest error rate among state-of-the-art NAS models, with 2.46% for CIFAR-10, 15.02% for CIFAR-100, and 23.8% for ImageNet dataset. Further, we implemented this image recognition architecture for other tasks, such as object detection, semantic segmentation, and keypoint detection tasks, and, in our experiments, EEEA-Net-C2 outperformed MobileNet-V3 on all of these various tasks. (The algorithm code is available at https://github.com/chakkritte/EEEA-Net).
Contamination may appear in various stages of hard disk manufacturing including the head gimbal assembly. Currently, the detection of contamination requires manual intervention. An image based automatic contamination detection strategy is therefore presented. After a preprocessing step, the contamination detection algorithm first detects potential areas of contamination using circle detection. Then, each of the contamination contenders are classified as either a contamination or noncontamination using a set of specific rules. The algorithm has been tested on 1,050 head gimbal assembly images of which 313 depicted contaminations. Our preliminary results yields an accuracy of 73.8% with a false negative rate of 34.8% and a false positive rate of 23.6%. Future work includes finetuning the contamination classification rules.
The detection of low quality solder joint quality in hard disk drive (HDD) manufacturing is a time consuming, error-prone and costly process that is often performed manually. This paper thus proposes two automated optical solder jet ball joint defect inspection methods for head gimbal assembly (HGA) production. The first method uses a Support Vector Machine (SVM) for fault detection and the second method uses vertical edge detection to identify solder ball and pad burning defects. The methods were tested with 5,530 HGA images, and their performance was compared to a Bayesian-based method. Experimental results show that the vertical edge detection method gave the best results, with an under reject rate of 0.75% and an over reject rate of 1.88%. The accuracy of the vertical edge detection method was 98.2%, which is higher than the accuracy of 89.9% for the Bayesian-based method, and 84.6% for the SVM-based method.
Index Terms-optical inspection; solder jet ball joint defect; vertical edge detection; HDD manufacture
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.