Deep Intelligence Learning Lab (DiL Lab) develops an intention prediction system that assists humans in intelligent living. DiL lab's research starts with an indoor localization system in which the study focuses on creating an advanced localization system with a 10cm localization error. The localization research addresses critical challenges such as cumulative error from IMU sensors, RSSI signal subject to refraction and attenuation, localization in complex environments, varying crowd densities, calibration errors, and multi-user localization. The lab also investigates human activity recognition (HAR) from daily life, mainly focusing on the sensor and camera-based HAR approaches. Our HAR research addresses the challenges, including the diversity of age, gender, and number of subjects, postural transitions, number of sensors and type of sensors, different body locations of wearable sensors or smartphones, missing values or labeling errors, similar postures, and datasets having complex activities, lack of ground truths, selection of appropriate datasets and selection of sensors. Emotion recognition also plays a significant role in the human intention prediction system. DiL lab's research focuses on facial expressions from a smartphone camera and emotion recognition from physiological signals (ECG and speech). The emotion recognition research addresses the main challenges, such as data augmentation, face occlusion, lighting issues, identifying facial features, recognizing unfinished emotions, racial differences, cultural differences in emotional expression, and identifying children's feelings. DiL lab research also uses videos to investigate the development of human pose (HPE) and eye-tracking systems for human intention prediction. The main challenges for human pose estimation are body pose variation, complicated background, and depth ambiguities. In the case of eye-tracking, our research introduces a novel technique continuously pursued to improve gaze detection precision and new ways to fully exploit eye data's potential. Our research outcomes are identifying human location, activity, expression, posture, and eye movements by implementing advanced deep learning models. The major applications of our human intention prediction system are ambient assisted living, intelligent and healthy living, healthcare, indoor navigation, and abnormal activity detection.