The rapidly changing landscapes of modern optimization problems require algorithms that can be adapted in real-time. This paper introduces an Adaptive Metaheuristic Framework (AMF) designed for dynamic environments. I...
详细信息
Inverter topologies for integrating a rooftop photovoltaic (PV) unit into a microgrid are becoming increasingly complex. This paper proposes a high-voltage boosting transformerless inverter (HVBTI) topology for enhanc...
详细信息
We present a miniature dual-wavelength fd-NIRS device for wearable real-time physiological monitoring. Based on a custom CMOS circuit, wearable device demonstrates accurate phase and amplitude measurements for modulat...
详细信息
We consider a two-player dynamic information design problem for a game played between a principal and a receiver on top of a Markovian system controlled by the receiver's actions. The principal strategically obtai...
详细信息
The contribution of energy efficiency covers a wide range of interests. Energy efficiency affects climate change mitigation goals, energy saving, health benefits, and productivity. The quick expansion of data traffic ...
详细信息
Visual question answering (VQA) aims at predicting an answer to a natural language question associated with an image. This work focuses on two important issues pertaining to VQA, which is a complex multimodal AI task:...
详细信息
the net of things (IoT) promises to revolutionize the way humans interact with their surroundings. Notwithstanding this promise, the deployment of IoT networks is hindered by using several challenges, particularly rel...
详细信息
Determining electromagnetic (EM) coupling to printed circuit boards (PCBs) is essential to finding potential EM susceptibilities early in the design process. For realistic PCB structures, analysis usually relies heavi...
详细信息
Identifying fruit disease manually is time-consuming, expertrequired,and expensive;thus, a computer-based automated system is widelyrequired. Fruit diseases affect not only the quality but also the *** a result, it is...
详细信息
Identifying fruit disease manually is time-consuming, expertrequired,and expensive;thus, a computer-based automated system is widelyrequired. Fruit diseases affect not only the quality but also the *** a result, it is possible to detect the disease early on and cure the fruitsusing computer-based techniques. However, computer-based methods faceseveral challenges, including low contrast, a lack of dataset for training amodel, and inappropriate feature extraction for final classification. In thispaper, we proposed an automated framework for detecting apple fruit leafdiseases usingCNNand a hybrid optimization algorithm. Data augmentationis performed initially to balance the selected apple dataset. After that, twopre-trained deep models are fine-tuning and trained using transfer ***, a fusion technique is proposed named Parallel Correlation Threshold(PCT). The fused feature vector is optimized in the next step using a hybridoptimization algorithm. The selected features are finally classified usingmachine learning algorithms. Four different experiments have been carriedout on the augmented Plant Village dataset and yielded the best accuracy of99.8%. The accuracy of the proposed framework is also compared to that ofseveral neural nets, and it outperforms them all.
暂无评论