• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.36 no.4
• /
• pp.294-299
• /
• 1991

Eight naked barley cultivars were grown in three production systems to select proper cultivars for dual production of forage and grain and to determine production system x cultivar interactions. In the forage systems. barley was seeded on September 27 and October 17, harvested for forage on November 1. December 10 and February 12 from the September 27 planting (forage system I) and on December 10 and February 12 from the October 17 seeding(forage system II). In the grain-only system, barley was planted on November 5. In forage system I, oven-dried forage yields of eight cultivars ranged 195 to 296kg/10a and Saessalbori and Naehanssalbori yielded significantly higher than the other cultivars recording 280 and 296kg/ l0a, respectively. In forage system II, oven-dried forage yields of eight cultivars ranged 106 to 143kg/10a showing no significant difference among cultivars. Production system x cultivar interactions were significant for lodging at maturity, powdery mildew rating, 1000 kernel weight and grain yield. Leaf area index and biomass at heading, no. of spikes per $m^2$, no. of kernels per spike and test weight were not affected by the production system. Forage utilization delayed heading by 3 days and reduced culm length, spike length, lodging due to heavy rain on May 5 and lodging at maturity except for Songhagbori and Naehanssalbori. Forage utilization did not significantly affect grain yield from the September 27 planting but reduced 9% from the October 17 planting. while Saessalbori and Hyangcheonggwa 1 yielded significantly less than in the grain-only system. Songhagbori appears to be a proper cultivar for dual production of forage and grain in Cheju, considering forage and grain yields. and lodging and powdery mildew resistance.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.1
• /
• pp.1-15
• /
• 2022

The MobileNetV3 is specially designed for mobile devices with limited memory and computing power. To reduce the network parameters and improve the network inference speed, a new lightweight network is proposed based on MobileNetV3. Firstly, to reduce the computation of residual blocks, a partial residual structure is designed by dividing the input feature maps into two parts. The designed partial residual structure is used to replace the residual block in MobileNetV3. Secondly, a dual-path feature extraction structure is designed to further reduce the computation of MobileNetV3. Different convolution kernel sizes are used in the two paths to extract feature maps with different sizes. Besides, a transition layer is also designed for fusing features to reduce the influence of the new structure on accuracy. The CIFAR-100 dataset and Image Net dataset are used to test the performance of the proposed partial residual structure. The ResNet based on the proposed partial residual structure has smaller parameters and FLOPs than the original ResNet. The performance of improved MobileNetV3 is tested on CIFAR-10, CIFAR-100 and ImageNet image classification task dataset. Comparing MobileNetV3, GhostNet and MobileNetV2, the improved MobileNetV3 has smaller parameters and FLOPs. Besides, the improved MobileNetV3 is also tested on CPU and Raspberry Pi. It is faster than other networks

• Korean Journal of Agricultural Science
• /
• v.12 no.2
• /
• pp.166-173
• /
• 1985

The multiple ear and tiller line (MET), which was selected from a local maize collection and has been selected for earliness by selfing for five generations, was compared its plant characteristics with four U.S. leading hybrids. 1. The general plant growth was a little retarded by poor soil fertility and structure. 2. The fresh weight of the MET line was about 97.8% of the U.S. check varieties. 3. The dry weight of the MET line was also about 94.5% of the check U.S. hybrids (1,695 kg / 10a.). 4. The tillering and eating habits of the MET line had contributed toward increasing fresh and dry weight of the MET line. 5. The average plant height of the MET line was much less than those of the check varieties, due to the selfing depression manifested by the MET line, (234.7cm. of the MET vs 250.7cm. of the check) 6. Apparent differences of disease resistance was observed among varieties studied. The degree of plants infected with the black stripe mosaic virus was over 20% for the check U.S. hybrids, while less than 5% of the MET line was infected with the disease. 7. The grain yield per 10 a of the MET line was far below the check hybrids due to the inbreeding depression of the MET line. However, the kernel number per unit area of the MET line with small size kernel was much greater than that of the check hybrids. The 100 kernel weight of the MET line was around 10 grams. 8. The average fresh and dry matter weight of the three ($sorghum{\times}sudan$ grass) varieties was very much alike with those of the MET line. 9. In conclusion, the use of the MET line for silage production may have dual adventages because of its low seed price and of its small size kernels, which may reduce the amount of seeds required for planting in an unit area.