Efficient Convolutional Neural Networks for Semiconductor Wafer Bin Map Classification.

The results obtained in the wafer test process are expressed as a wafer map and contain important information indicating whether each chip on the wafer is functioning normally. The defect patterns shown on the wafer map provide information about the process and equipment in which the defect occurred, but automating pattern classification is difficult to apply to actual manufacturing sites unless processing speed and resource efficiency are supported. The purpose of this study was to classify these defect patterns with a small amount of resources and time. To this end, we explored an efficient convolutional neural network model that can incorporate three properties: (1) state-of-the-art performances, (2) less resource usage, and (3) faster processing time. In this study, we dealt with classifying nine types of frequently found defect patterns: center, donut, edge-location, edge-ring, location, random, scratch, near-full type, and None type using open dataset WM-811K. We compared classification performance, resource usage, and processing time using EfficientNetV2, ShuffleNetV2, MobileNetV2 and MobileNetV3, which are the smallest and latest light-weight convolutional neural network models. As a result, the MobileNetV3-based wafer map pattern classifier uses 7.5 times fewer parameters than ResNet, and the training speed is 7.2 times and the inference speed is 4.9 times faster, while the accuracy is 98% and the F1 score is 89.5%, achieving the same level. Therefore, it can be proved that it can be used as a wafer map classification model without high-performance hardware in an actual manufacturing system.

Susceptibility of the Cigarette Beetle Lasioderma serricorne (Fabricius) to Phosphine, Ethyl Formate and Their Combination, and the Sorption and Desorption of Fumigants on Cured Tobacco Leaves.

The susceptibility of Lasioderma serricorne to phosphine (PH3), ethyl formate (EF) and their combination (PH3 + EF) was evaluated in this study. Eggs, larvae, pupae and adults were subjected to treatment with fumigants to determine the 90% lethal concentration time (LCt90) values. Treatment with PH3 for 20 h resulted in LCt90 values of 1.15, 1.39, 14.97 and 1.78 mg h/L while treatment with EF resulted in values of 157.96, 187.75, 126.06 and 83.10 mg h/L, respectively. By contrast, the combination of PH3 + EF resulted in LCt90 values of 36.05, 44.41, 187.17 and 35.12 mg h/L after 4 h. These results show that, through treatment with PH3 + EF, control can be achieved at lower concentrations than for treatment with EF alone and at lower exposure times than for treatment with PH3 alone. The sorption rates of the fumigants on cured tobacco leaves were determined for filling ratios of 2.5%, 5.0% and 10.0% (w/v). Cured tobacco leaves were treated with either 2 mg/L PH3, 114 mg/L EF or 0.5 mg/L PH3 + 109 mg/L EF. Treatment with PH3 showed sorption rates of 0.0%, 7.1% and 14.3%. EF, however, showed higher sorption rates of 64.9%, 68.5% and 75.5%, respectively, for the indicated filling ratios. When PH3 and EF were combined, the sorption rate of PH3 was 0.0%, while the sorption rates of EF were lower (9.1%, 12.0% and 23.2%) than treatment with only EF. EF required a ventilation time of longer than 22 h to desorb from cured tobacco leaves. Therefore, PH3 + EF can effectively control L. serricorne in cured tobacco leaves, with sufficient ventilation time required after treatment for the safety of workers.

Insecticidal Effects of Fumigants (EF, MB, and PH3) towards Phosphine-Susceptible and -Resistant Sitophilus oryzae (Coleoptera: Curculionidae).

This study was conducted to evaluate the insecticidal effects of ethyl formate (EF), methyl bromide (MB), and phosphine (PH3) fumigants against PH3-susceptible and -resistant strains of the rice weevil (Sitophilus oryzae), a major rice pest. The highest lethal concentration time 50 (LCt50) values of the PH3-susceptible strains were 255.797, 21.104, and 6.171 mg h L-1 for EF, MB, and PH3, respectively, at pupal stage. The highest LCt50 values of the PH3-resistant strains were 149.028 and 48.170 mg h L-1 for EF and PH3, respectively, at late larval stage, and 43.520 mg h L-1 for MB at pupal stage. In comparison to the susceptible strains, the PH3-resistant strain collected in South Korea had a resistance level 4 to 56 times higher. Use of the major quarantine fumigants EF, MB, and PH3 indicated the existence of PH3-resistant rice weevils in South Korea for the first time.

Prevalence of Bacillus cereus bacteriophages in fermented foods and characterization of phage JBP901.

Bacillus cereus is widely distributed in the environment and is responsible for foodborne illnesses, causing diarrhea and vomiting. In this study, the prevalence of B. cereus bacteriophages in Korean fermented foods was studied using 6 reference strains, and JBP901 was further characterized. In a total of 47 samples, 19 (40.4%) contained more than one kind of B. cereus phage. Fourteen phages were purified and determined for host lysis profile. Among them, JBP901 was specific to B. cereus group strains (forming plaques in 13 of 14 isolates) and not able to lyse other tested Gram-positive and Gram-negative strains including Bacillus subtilis. Transmission electron microscopy analysis suggested that JBP901 is a member of the family Myoviridae. Restriction enzyme digestion pattern analysis and one-step growth study indicated a genome size of approximately 150 kbp and a burst size of 100. Furthermore, bacterial lysis studies revealed that JBP901 could be used to control the growth of B. cereus in liquid culture and in food. This is the first report to examine the prevalence of phages in fermented foods in Korea, and suggests that fermented foods are useful sources of B. cereus phage isolation; moreover, B. cereus phages could provide an additional tool to tackle problems associated with B. cereus contamination.