Human Activity Recognition (HAR) is an important area of research due to its applications in health monitoring, elderly care, and personal fitness tracking. The challenge is deploying efficient and accurate HAR systems on resource-constrained embedded devices, which require low power consumption and processing efficiency. This work optimizes a Convolutional Neural Network (CNN) model for HAR, targeting resource-constrained processors. The goal is to balance accuracy, performance, and power consumption for real-world deployment in wearable devices. Key contributions include introducing an Extended 1D CNN model that enhances temporal awareness and accuracy without the overhead of floating-point computations, evaluating and applying quantization methods to minimize model size with minimal accuracy loss, and assessing the model's performance on a RISC-V processor. Results show an accuracy increase from 74% to 87.2%. Memory optimization using Lookup Table (LUT) quantization reduces the memory required for model parameters by 57%. This research underscores the potential for advanced neural network models on low-power RISC-V processors in real-time HAR, with significant implications for health monitoring and smart environments.