Early detection of plant diseases is essential for effective crop disease management to prevent yield loss. In this study, we developed a methodology for classifying diseases in rice leaves using four deep learning models and a dataset with 2658 images of healthy and diseased rice leaves. Four models, namely LeafNet, Modified LeafNet, MobileNetV2, and Xception, were compared. The Modified LeafNet model involved updates to LeafNet’s architectural parameters, whereas transfer learning techniques were applied to the MobileNetV2 and Xception pretrained models. The optimal hyperparameters for training were determined by considering several factors such as batch size, data augmentation, learning rate, and optimizers. The Modified LeafNet model achieved the highest accuracies of 97.44% and 87.76% for the validation and testing datasets, respectively. In comparison, LeafNet obtained 88.92% and 71.84%, Xception obtained 88.64% and 71.95%, and MobileNetV2 obtained 82.10% and 67.68% for the validation and test accuracies on the same datasets, respectively. This study contributes to the development of automated disease classification systems for rice leaves, thereby leading to increased agricultural productivity and sustainability.

Several studies on micro-expression recognition have contributed mainly to accuracy improvement. However, the computational complexity receives lesser attention comparatively and therefore increases the cost of micro-expression recognition for real-time application. In addition, majority of the existing approaches required at least two frames (i.e., onset and apex frames) to compute features of every sample. This paper puts forward new facial graph features based on 68-point landmarks using Facial Action Coding System (FACS). The proposed feature extraction technique (FACS-based graph features) utilizes facial landmark points to compute graph for different Action Units (AUs), where the measured distance and gradient of every segment within an AU graph is presented as feature. Moreover, the proposed technique processes ME recognition based on single input frame sample. Results indicate that the proposed FACS-baed graph features achieve up to 87.33% of recognition accuracy with F1-score of 0.87 using leave one subject out cross-validation on SAMM datasets. Besides, the proposed technique computes features at the speed of 2 ms per sample on Xeon Processor E5-2650 machine.

This paper presents an indoor positioning system (IPS) and motion tracking system for the elderly, who are staying alone. The system is able to track the elderly location based on the room that they are currently situated and recognizes whether the elderly is moving or sitting. A smart wearable device featuring MPU6050 motion sensor, BLE beacon and Raspberry Pi Zero W is developed. The sensory data from MPU6050 will be analyzed to study the type of motion and publish to MQTT server. For location tracking, multiple Raspberry Pi 3 are used as BLE beacon scanners to scan for the beacon and publish its signal strength to an MQTT server. The datasets for the indoor location and motion type have also been developed in this research work. Weka, a machine learning software is used to evaluate and determine the most suitable classifier. The machine learning technique was implemented in Python programming language using scikit-learn, matplotlib, numpy and SciPy packages. The results are then pushed to an Internet of Things (IoT) platform for remote monitoring. The accuracy rate is 99% for the location tracking and 99.97% for motion recognition.

Behavioral changes in daily home activities may be linked with health problems. Therefore, anomaly detection on sequence pattern of home activities is important for healthcare monitoring. In this paper, an anomaly detection method based on Long Short-Term Memory (LSTM) neural network is proposed to detect anomalies on sequence pattern of home activities. A comparison study of LSTM and Hidden Markov Model (HMM) was conducted to evaluate their performance under different training set size and model's hyperparameters. The experimental results demonstrated that LSTM is comparable to HMM in detecting anomalies on sequence pattern of home activities. The test accuracies of the best LSTM and HMM models are both 87.50%.

An Internet of Things (IoT) FPGA-based home hub to automate control operations in a home environment was designed and built. The proposed system uses an FPGA home hub as its local analytic engine with an IoT platform to store the sensory data. The FPGA was programmed in Verilog HDL using Quartus II provided by Altera. The WiFi capability of the FPGA was extended through an ESP8266 chip to ease the interfacing with various sensors connected to it. The system can be configured and monitored through a web application coded in JavaScript. Various test cases were carried out on the implemented system at Multimedia University (MMU) Digital Home Lab. The results verified the functionality of the system in triggering real home appliances (i.e. air conditioning unit and lighting) based on multiple sensor nodes without conflicting each other. The ability to allow user to configure the control rules based on the sensory data via web interface hosted using ThingSpeak Plugins is also presented and demonstrated in this project. The base design is utilizing Altera Cyclone IV EP4CE22F17C6N FPGA with 153 I/O pins, which is highly scalable and adaptable to the requirements of home environments. This shows promising application of FPGA in supporting scalable IoT home automation system.

The rapid growth of the ageing population has increased the attention paid towards the provision of retirement village (RV). RV is a new form of housing on the ageing landscape that provides a full range of care from independent to assisted living to meet the diverse needs of the elderly. Numerous studies conducted in the past examined factors that influence elderly intention to relocate to the RV and the intention to adopt the smart home technology. The current study focused on the intention of elderly to relocate to the smart retirement village (SRV) (i.e., RV with the smart home technology). A customer-driven open innovation survey was conducted to solicit opinions from the elderlies. This paper applied the Theory of Planned Behavior to examine such intention among the elderly in Malaysia. The types of smart home technology preferred by the elderly at the SRV were determined. The results from the multiple regression analysis (N = 305) indicated that attitudes, social norm, and perceived behavioural control contributed significantly to the elderly intention to relocate to the SRV. Further investigation revealed that 83% of the elderly choose the smart emergency contact system as the most important smart home technology to be included.