 Hi, I'm Shen. I'm from the University of Virginia. The work I present today is deciphering human emotion communication through slight changes of touch. In our social activities, we often communicate distinct emotional sentiments through nuance of the touch. For instance, we might generally hold one's arm to offer a sense of calm, yet intensively hold one's arm to express anger or anxiety. As it indicates, distinct sentiments may be shaped by the subtle in one's touch delivery. However, such contact details are still not fully understood due to the difficulty in quantifying naturalistic human touch. Among tracking techniques commonly used for human touch, motion trackers can only capture the position or skeleton information while pressure mass sensors block the direct contact. Therefore, we developed a 3D visual tracking system in this work that captures the toucher's hand and touch receiver's forearm at a 3D point called an mesh. We then derived the contact attribute through in touch, including contact area, indentation depth, spatial hand velocity, three velocity components, and the overall contact duration. In the psychophysical experiments, the toucher was instructed to express emotions by touching the receiver's forearm with preferred gestures. After the touch, the receiver selected a message to identify from the contact and rate the balance and arousal accordingly. Based on experiments, we found that even the same touch gesture was used, people can still slightly but significantly vary the contact attributes to express totally different emotions. For example, people can use stroking gesture to express calm, sympathy, and happiness, while happiness is typically delivered with a much higher stroking velocity. We further summarized the contact profile of each emotional message in the terms of contact attributes. The importance of each attribute in discriminating those messages were also derived, which could aid in the design and optimization of the digital social touch. In addition to discrete messages, we also found that contact attributes correlate with the emotional ratings of balance and arousal, especially for the stroking gesture. An inverted U-shaped curve was fitted between the balance rating and the velocity. This correlation is well aligned with the trend widely reported for the pleasant rating and the c-tail-tail-faring activity induced by controlled brush stimuli. Furthermore, arousal rating exhibits similar positive correlation with stroking velocity as other subtypes of peripheral afferents such as SA1 and SA2, which might hint at their role in signaling arousal percents.