Fine-grained dog breed classification presents significant challenges due to subtle inter-class differences,pose variations,and intra-class diversity.To address these complexities and limitations of traditional handcr...Fine-grained dog breed classification presents significant challenges due to subtle inter-class differences,pose variations,and intra-class diversity.To address these complexities and limitations of traditional handcrafted approaches,a novel and efficient two-stageDeep Learning(DL)framework tailored for robust fine-grained classification is proposed.In the first stage,a lightweight object detector,YOLO v8N(You Only Look Once Version 8 Nano),is finetuned to localize both the head and full body of the dog from each image.In the second stage,a dual-stream Vision Transformer(ViT)architecture independently processes the detected head and body regions,enabling the extraction of region-specific,complementary features.This dual-path approach improves feature discriminability by capturing localized cues that are vital for distinguishing visually similar breeds.The proposed framework introduces several key innovations:(1)a modular and lightweight head–body detection pipeline that balances accuracy with computational efficiency,(2)a region-awareViT model that leverages spatial attention for enhanced fine-grained recognition,and(3)a training scheme incorporating advanced augmentations and structured supervision to maximize generalization.These contributions collectively enhancemodel performancewhilemaintaining deployment efficiency.Extensive experiments conducted on the Tsinghua Dogs dataset validate the effectiveness of the approach.The model achieves an accuracy of 90.04%,outperforming existing State-of-the-Art(SOTA)methods across all key evaluation metrics.Furthermore,statistical significance testing confirms the robustness of the observed improvements over multiple baselines.The proposed method presents an effective solution for breed recognition tasks and shows strong potential for broader applications,including pet surveillance,veterinary diagnostics,and cross-species classification.Notably,it achieved an accuracy of 96.85% on the Oxford-IIIT Pet dataset,demonstrating its robustness across different species and breeds.展开更多
Fused deposition modelling (FDM) is one of rapid prototyping (RP) technologies which uses an additive fabrication approach.Each commercially available FDM model has different types of process parameters for different ...Fused deposition modelling (FDM) is one of rapid prototyping (RP) technologies which uses an additive fabrication approach.Each commercially available FDM model has different types of process parameters for different applications.Some of the desired parts require excellent surface finish as well as good tolerance.The most common parameters requiring setup are the raster angle,tool path,slice thickness,build orientation,and deposition speed.The purpose of this paper is to discuss the process parameters of FDM Prodigy Plus (Stratasys,Inc.,Eden Prairie,MN,USA).Various selected parameters were tested and the optimum condition was proposed.The quality of the parts produced was accessed in terms of dimensional accuracy and surface finish.The optimum parameters obtained were then applied in the fabrication of the master pattern prior to silicone rubber moulding (SRM).These parameters would reduce the post processing time.The dimensional accuracy and surface roughness were analyzed using coordinate measuring machine (CMM) and surface roughness tester,respectively.Based on this study,the recommended parameters will improve the quality of the FDM parts produced in terms of dimensional accuracy and surface roughness for the application of SRM.展开更多
Sun radiation is known as one of the cleanest energy available and free maintenance on the earth to support the extensive demand for electrical power requirement for large and small scale. This energy is known as most...Sun radiation is known as one of the cleanest energy available and free maintenance on the earth to support the extensive demand for electrical power requirement for large and small scale. This energy is known as most promising renewable energy available during the whole year. This fact has shown a very strong interest among the researchers, industries and consumers to develop technology and system that can harvest solar energy to electrical energy. A lot of researches have been done to explore the capability of sun as natural energy source which is capable to provide huge alternative support to the increasing demand of electrical power in society. All the research findings have been adaptively combined to design, develop and structure as a reliable finished product for end users implementation. The fast improving research in photovoltaic application has proven that photovoltaic renewable energy is considered as one of the most promising source to support the increasing demand in electrical utilization. Advantages implementing the photovoltaic technology will benefit during the fuel price increment and be insufficient of the non-renewable energy to support the increasing demand from the society. Looking into this factor, many industries have developed the electrical photovoltaic system in large, medium and small scale for implementation. Electricity is becoming an essential requirement because of the increasing demand for electrical power from the society. Due to the increasing demand for electrical power, many researches are conducted to support the conventional way of generating electricity. This research has explored many other available abundant renewable energy resources such as solar, wind and thermal which are largely available to support the increasing demand. In this research, researchers discovered that the solar energy is the main contributor to support the increasing demand for electricity. The research outcome has helped the industries to develop solar power photovoltaic systems. This paper discusses the approaches to design and develop a real-time 500 Watt DC to AC Solar Power System. The system capability to store the harvested energy is discussed in this paper.展开更多
基金supported by the faculty research fund of Sejong University in 2023,the National Research Foundation ofKorea(NRF)grant funded by theKorea government(MSIT)(RS-2025-00518960)Institute of Information&Communications Technology Planning&Evaluation(IITP)under the metaverse support program to nurture the best talents(IITP-2025-RS-2023-00254529)grant funded by the Korea government(MSIT).
文摘Fine-grained dog breed classification presents significant challenges due to subtle inter-class differences,pose variations,and intra-class diversity.To address these complexities and limitations of traditional handcrafted approaches,a novel and efficient two-stageDeep Learning(DL)framework tailored for robust fine-grained classification is proposed.In the first stage,a lightweight object detector,YOLO v8N(You Only Look Once Version 8 Nano),is finetuned to localize both the head and full body of the dog from each image.In the second stage,a dual-stream Vision Transformer(ViT)architecture independently processes the detected head and body regions,enabling the extraction of region-specific,complementary features.This dual-path approach improves feature discriminability by capturing localized cues that are vital for distinguishing visually similar breeds.The proposed framework introduces several key innovations:(1)a modular and lightweight head–body detection pipeline that balances accuracy with computational efficiency,(2)a region-awareViT model that leverages spatial attention for enhanced fine-grained recognition,and(3)a training scheme incorporating advanced augmentations and structured supervision to maximize generalization.These contributions collectively enhancemodel performancewhilemaintaining deployment efficiency.Extensive experiments conducted on the Tsinghua Dogs dataset validate the effectiveness of the approach.The model achieves an accuracy of 90.04%,outperforming existing State-of-the-Art(SOTA)methods across all key evaluation metrics.Furthermore,statistical significance testing confirms the robustness of the observed improvements over multiple baselines.The proposed method presents an effective solution for breed recognition tasks and shows strong potential for broader applications,including pet surveillance,veterinary diagnostics,and cross-species classification.Notably,it achieved an accuracy of 96.85% on the Oxford-IIIT Pet dataset,demonstrating its robustness across different species and breeds.
文摘Fused deposition modelling (FDM) is one of rapid prototyping (RP) technologies which uses an additive fabrication approach.Each commercially available FDM model has different types of process parameters for different applications.Some of the desired parts require excellent surface finish as well as good tolerance.The most common parameters requiring setup are the raster angle,tool path,slice thickness,build orientation,and deposition speed.The purpose of this paper is to discuss the process parameters of FDM Prodigy Plus (Stratasys,Inc.,Eden Prairie,MN,USA).Various selected parameters were tested and the optimum condition was proposed.The quality of the parts produced was accessed in terms of dimensional accuracy and surface finish.The optimum parameters obtained were then applied in the fabrication of the master pattern prior to silicone rubber moulding (SRM).These parameters would reduce the post processing time.The dimensional accuracy and surface roughness were analyzed using coordinate measuring machine (CMM) and surface roughness tester,respectively.Based on this study,the recommended parameters will improve the quality of the FDM parts produced in terms of dimensional accuracy and surface roughness for the application of SRM.
文摘Sun radiation is known as one of the cleanest energy available and free maintenance on the earth to support the extensive demand for electrical power requirement for large and small scale. This energy is known as most promising renewable energy available during the whole year. This fact has shown a very strong interest among the researchers, industries and consumers to develop technology and system that can harvest solar energy to electrical energy. A lot of researches have been done to explore the capability of sun as natural energy source which is capable to provide huge alternative support to the increasing demand of electrical power in society. All the research findings have been adaptively combined to design, develop and structure as a reliable finished product for end users implementation. The fast improving research in photovoltaic application has proven that photovoltaic renewable energy is considered as one of the most promising source to support the increasing demand in electrical utilization. Advantages implementing the photovoltaic technology will benefit during the fuel price increment and be insufficient of the non-renewable energy to support the increasing demand from the society. Looking into this factor, many industries have developed the electrical photovoltaic system in large, medium and small scale for implementation. Electricity is becoming an essential requirement because of the increasing demand for electrical power from the society. Due to the increasing demand for electrical power, many researches are conducted to support the conventional way of generating electricity. This research has explored many other available abundant renewable energy resources such as solar, wind and thermal which are largely available to support the increasing demand. In this research, researchers discovered that the solar energy is the main contributor to support the increasing demand for electricity. The research outcome has helped the industries to develop solar power photovoltaic systems. This paper discusses the approaches to design and develop a real-time 500 Watt DC to AC Solar Power System. The system capability to store the harvested energy is discussed in this paper.
