Hand detection is a fundamental step for many hand-related computer vision tasks, such as gesture recognition, hand pose estimation, hand sign language translation, and so on. However, robustly detecting hands is a challenging task because of drastic changes in appearance based on finger articulation and changes in lighting conditions, camera angles, backgrounds, skin colors in images. Adult hand detection has been previously addressed; however, recent hand detection systems perform poorly for infants. This thesis presents a robust end- to-end infant hand detection and tracking algorithm based on Faster RCNN for hand detection, fine-tuned on a custom dataset of infant’s hands. We then use the generated infant hand tracks to develop other downstream applications to help psychologists analyze fine-grained object-hand interactions to study how infants afford everyday objects.