This article presents a low-power, high-speed smart vision sensor for motion detection (MD) that realizes in-pixel frame-difference (FD) operation using a global shutter mechanism. A ping-pong pulse-width-modulation (PWM) pixel is proposed to achieve the consecutive event frame report with a balanced signal transfer function of successive FD operations. Three operating modes were implemented for varied application scenarios, such as image capture (IC) mode to capture a raw image, FD mode for MD, and saliency detection (SD) mode for low-resolution sub-block event counting. A 0.8 V 64 <inline-formula> <tex-math notation="LaTeX">\times </tex-math></inline-formula> 64 vision sensor prototype was fabricated and verified in TSMC 0.18-<inline-formula> <tex-math notation="LaTeX">\mu \text{m} </tex-math></inline-formula> standard CMOS technology. In IC mode, it consumed 71.2 <inline-formula> <tex-math notation="LaTeX">\mu \text{W} </tex-math></inline-formula>@360fps with an achieved iFoM of 48.3 <inline-formula> <tex-math notation="LaTeX">\text {pJ/pixel}\cdot \text {frame} </tex-math></inline-formula>. In FD mode, it consumed 74.4 <inline-formula> <tex-math notation="LaTeX">\mu \text{W} </tex-math></inline-formula>@510fps with full-resolution (64 <inline-formula> <tex-math notation="LaTeX">\times </tex-math></inline-formula> 64) event reporting and achieved iFoMs of 35.6 <inline-formula> <tex-math notation="LaTeX">\text {pJ/pixel}\cdot \text {frame} </tex-math></inline-formula>. In SD mode, it consumed 121.6 <inline-formula> <tex-math notation="LaTeX">\mu \text{W} </tex-math></inline-formula>@890fps with block-level (8 <inline-formula> <tex-math notation="LaTeX">\times </tex-math></inline-formula> 8) saliency reporting and achieved iFoMs of 2.1 <inline-formula> <tex-math notation="LaTeX">\text {nJ/block}\cdot \text {frame} </tex-math></inline-formula>.