• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Journal of Korean Society of Environmental Engineers
• /
• v.28 no.8
• /
• pp.866-871
• /
• 2006

The nutrient recovery in phosphate crystallization process was investigated by using laboratory scale uptlow reactors, adopting sequencing batch type configuration. The industrial waste lime was used as potential cation source with magnesium salt($MgCl_2$) as control. The research was focused on its successful application in a novel integrated sludge treatment process, which is comprised of a high performance fermenter followed by a crystallization reactor. In the struvite precipitation test using synthetic wastewater first, which has the similar characteristics with the real fermentation effluent, the considerable nutrient removal(about 60%) in both ammonia and phosphate was observed within $0.5{\sim}1$ hr of retention time. The results also revealed that a minor amount(<5%) of ammonia stripping naturally occurred due to the alkaline(pH 9) characteristic in feed substrate. Stripping of $CO_2$ by air did not increase the struvite precipitation rate but it led to increased ammonia removal. In the second experiment using the fermentation effluent, the optimal dosage of magnesium salt for struvite precipitation was 0.86 g Mg $g^{-1}$ P, similar to the mass ratio of the struvite. The optimal dosage of waste lime was 0.3 g $L^{-1}$, resulting in 80% of $NH_4-N$ and 41% of $PO_4-P$ removal, at about 3 hrs of retention time. In the microscopic analysis, amorphous crystals were mainly observed in the settled solids with waste lime but prism-like crystals were observed with magnesium salt. Based on mass balance analysis for an integrated sludge treatment process(fermenter followed by crystallization reactor) for full-scale application(treatment capacity Q=158,880 $m^3\;d^{-1}$), nutrient recycle loading from the crystallization reactor effluent to the main liquid stream would be significantly reduced(0.13 g N and 0.19 g P per $m^3$ of wastewater, respectively). The results of the experiment reveal therefore that the reuse of waste lime, already an industrial waste, in a nutrient recovery system has various advantages such as higher economical benefits and sustainable treatment of the industrial waste.

• Korean Journal of Poultry Science
• /
• v.8 no.2
• /
• pp.55-68
• /
• 1981

Calcium and phosphorus are not only indispensable for the bone formation and body fluids equilibrium but also are major components of egg shell. It is nutritionally important, therefore, to investigate the metabolism of calcium and phosphorus and to search for optimum requirement of calcium and phosphorus and the availability of various sources of calcium an4 phosphorus by poultry. An attempt was made to review the nutrition of calcium and phosphorus in poultry diets. 1, Calcium and phosphorus have great interrelationship with vitamin D in their metabolisms. 2. Most of the plant-origin phosphorus are existing in phytic form and it leads to low availability when used in poultry rations, although calcium and phosphorus present in animal-origin or mineral supplements are highly available in general. 3. Calcium and phosphorus requirement from existing information indicated that 1.0% calcium and 0.7% phosphorus for broiler and egg-type chicks, and 3.5% calcium and 0.4% phosphorus for laying hen. 4. It has been recommended that calcium and phosphorus level should be increased when the feed intake was decreased or when the egg Production rate was higher or when the hens are old. 5. Mono-, ci-, tri-, calcium phosphate, calcium carbonate, bone meal, limestone and oyster shell u the most readily available among various sources of calcium phosphorus supplements. Soft rock phosphate, deflourinated phosphate and gypsum are somewhat inferior to the previous ones in bioavailability. 6. The effect of particle size of calcium supplements on egg shell quality and egg production rate is not yet clearly defined but recent works showed that oyster shell is more available when it was coarse and limestone is more available when it was fine in panicle. size. 7. Present data indicated that mixed feeding of oyster shell and limestone is superior to the single feeding of each on laying performance. 8. Significant interaction between phosphorus and sodium was observed, that is, excessive sodium decreased egg production in layer and body weight growth in broiler in the low phosphorus diets but increased them in the high phosphorus diets.

• ANNALS OF ANIMAL RESOURCE SCIENCES
• /
• v.29 no.4
• /
• pp.172-182
• /
• 2018

The present study was undertaken to examine if the carcass and pork quality of finishing pigs reared on a low plane of nutrition (LPN) could be improved compared with that of the pigs finished on a high plane of nutrition (HPN). Sixty-eight crossbred (LYD) barrows and 68 LYD gilts weighing approximately 50 kg were fed a diet containing 3.54 Mcal DE/kg with 1.00% lysine (HPN) or 3.02 Mcal DE/kg with 0.68% lysine (LPN) in eight pens up to approximately 120 kg and slaughtered. The belly, loin, ham, and Boston butt were cut out from a total of 20 carcasses, after which physicochemical and sensory quality attributes of the belly and the representative muscle of each of the loin, ham, and Boston butt were evaluated. The ADG, gain:feed ratio, and backfat thickness were less for LPN than for HPN (p<0.05). The cooking loss, hardness, and chewiness values for the Boston butt were less for LPN vs. HPN. In sensory evaluation for fresh meat (muscle), the subjective quality scores were greater for LPN vs. HPN in color, marbling, and acceptability for the loin, the muscle:fat balance score for the belly tending to be greater for LPN (p<0.10). In addition, LPN was superior to HPN in the flavor and juiciness in sensory evaluation for cooked ham. In conclusion, the present results suggest that the carcass and pork quality of finishing pigs could be improved with reduced growth performance by using LPN.

• Journal of Life Science
• /
• v.32 no.8
• /
• pp.601-610
• /
• 2022

P. tenebrifer (PT) belongs to the Diptera order and Stratiomyidae family. Recently, insect industry have been focused as food, animal feed and environmental advantages. γ-aminobutyric acid (GABA) and melatonin have been associated with regulating sleep and depression. GABA is the primary inhibitory neurotransmitter and is synthesized via biotransformation of monosodium glutamate (MSG) to GABA by lactic acid bacteria. In this study, we first used a GABA-enhanced PT extract, wherein GABA was enhanced by feeding MSG to PT. The underlying mechanisms preventing stress and insomnia were investigated in a corticosterone (CORT)-induced endoplasmic reticulum (ER) stress and chronic restraint stress (CRS)-exposed mouse model, as well as in pentobarbital (45 mg/kg)-induced sleep behaviors in mice. In the present study, the GABA peak was detected in high-performance liquid chromatography-evaporative light scattering detector (HPLC-ELSD) analysis and showed in Ptecticus tenebrifer water extract (PTW) but not in non-PTW extract. The results showed that PTW and Ptecticus tenebrifer with 70% ethanol extract (PTE) exerted neuroprotective effects by protecting against CORT-induced downregulation of phosphorylated extracellular signal-regulated kinase 1/2 (ERK1/2) and cAMP-response element binding protein (CREB) expression. In addition, PTW (300 mg/kg) significantly reduced CORT levels in CRS-exposed mice. Furthermore, PTW (100 and 300 mg/kg) significantly reduced sleep latency and increased total sleep duration in pentobarbital (45 mg/kg)-induced sleeping behaviors, which was related to serum melatonin levels. In conclusion, our results suggest that PTW exerts anti-stress and sleep-enhancing effects by regulating serum CORT and melatonin levels.

• Proceedings of the Korean Society for Food Science of Animal Resources Conference
• /
• 1998.05a
• /
• pp.37-54
• /
• 1998

This study was conducted to investigate the effects of mugwort extracts on the blood ethanol concentration, liver function and low level of cadmuim(Cd) in rats. The effects of mugwort extracts on the blood ethanol concentration was studied in Sprague-Dawley rats (10 weeks old) administered p.o. with 25% ethanol (5g/1kg body weight) and then injected with mugwort extracts (at the 2% levels of daily feed consumption compared with the concentration of catechins level in mugwort extracts) in caudal vein. SD rats were divided into five groups : control group (CON-E, only ethanol and 0.85% saline sol'n treated instead of each extracts), water extracts of mugwort treated to the control (MDW-E), ethanol extracts of mugwort treated to the control (POH-E). And then rat plasma of each time (0hr, 1hr, 2hr, 3hr) was investigated ethanol concentration by gas chromatography. Another rats were measured at the time of 0 and 5hr for the test of GOD(Glutamic Oxaloacetic Transaminase) and GPT(Glutamic Pyruvic Transaminase). Components of each extracts were analyzed by using high performance liquid chromatography. The effects of mugwort extracts on the liver function were studied in culture of rat hepatocyte composed of three groups : Control group and two groups treated with each extracts (1% & 2% MDW, 1% & 2% MOH). Condition of rat hepatocytes cultured for 36hr at $37^{\circ}C$(5% $CO_2$ incubator), number of cells, GOT and GPT activity were investigated. The results obtained were summarized as follows ; 1. Catechins level of mugwort extracts was $8{\sim}10mg/100g(MDW)$, $3{\sim}4mg/100g(MOH)$ 2. The contents of (-)-Epigallocatechin was high in MDW 3. The effects of mugwort extracts on the blood ethanol concentration were as follows; 1) The order in ethanol degradation efficiency was MDW-E > MOH-E > CON-E. 2) Ethanol concentration significantly decreased (p<0.05) in MDW-E and MOH-E. 4. The effects of mugwort extracts on the liver function were as follows; (rat hepatocytes cultured for 36hr at $37^{\circ}C$) 1) Cells condition of MDW-L was better than other groups. 2) The order in number of cells (rat hepatocytes) was 2% MDW-L >1% MDW-L >1% MOH-L > Con-L > 2% MOH-L 5. Cd treatment increased concentrations of hepatic GSH level, and decreased GOT activity in plasma. Therefore, this results suggest that the effects of mugwort extracts may an important rols in degradation ethanol and recovery liver function in body. Also, Mugwort extracts may modify the toxicities of Cd in Cd-treated rats and play an important roles in preventing the liver from various toxicants including Cd in Cd treated rats.

• Journal of Intelligence and Information Systems
• /
• v.24 no.4
• /
• pp.137-154
• /
• 2018

Animal infectious diseases, such as avian influenza and foot and mouth disease, occur almost every year and cause huge economic and social damage to the country. In order to prevent this, the anti-quarantine authorities have tried various human and material endeavors, but the infectious diseases have continued to occur. Avian influenza is known to be developed in 1878 and it rose as a national issue due to its high lethality. Food and mouth disease is considered as most critical animal infectious disease internationally. In a nation where this disease has not been spread, food and mouth disease is recognized as economic disease or political disease because it restricts international trade by making it complex to import processed and non-processed live stock, and also quarantine is costly. In a society where whole nation is connected by zone of life, there is no way to prevent the spread of infectious disease fully. Hence, there is a need to be aware of occurrence of the disease and to take action before it is distributed. Epidemiological investigation on definite diagnosis target is implemented and measures are taken to prevent the spread of disease according to the investigation results, simultaneously with the confirmation of both human infectious disease and animal infectious disease. The foundation of epidemiological investigation is figuring out to where one has been, and whom he or she has met. In a data perspective, this can be defined as an action taken to predict the cause of disease outbreak, outbreak location, and future infection, by collecting and analyzing geographic data and relation data. Recently, an attempt has been made to develop a prediction model of infectious disease by using Big Data and deep learning technology, but there is no active research on model building studies and case reports. KT and the Ministry of Science and ICT have been carrying out big data projects since 2014 as part of national R &D projects to analyze and predict the route of livestock related vehicles. To prevent animal infectious diseases, the researchers first developed a prediction model based on a regression analysis using vehicle movement data. After that, more accurate prediction model was constructed using machine learning algorithms such as Logistic Regression, Lasso, Support Vector Machine and Random Forest. In particular, the prediction model for 2017 added the risk of diffusion to the facilities, and the performance of the model was improved by considering the hyper-parameters of the modeling in various ways. Confusion Matrix and ROC Curve show that the model constructed in 2017 is superior to the machine learning model. The difference between the2016 model and the 2017 model is that visiting information on facilities such as feed factory and slaughter house, and information on bird livestock, which was limited to chicken and duck but now expanded to goose and quail, has been used for analysis in the later model. In addition, an explanation of the results was added to help the authorities in making decisions and to establish a basis for persuading stakeholders in 2017. This study reports an animal infectious disease prevention system which is constructed on the basis of hazardous vehicle movement, farm and environment Big Data. The significance of this study is that it describes the evolution process of the prediction model using Big Data which is used in the field and the model is expected to be more complete if the form of viruses is put into consideration. This will contribute to data utilization and analysis model development in related field. In addition, we expect that the system constructed in this study will provide more preventive and effective prevention.