Hand Gesture Gesture Detection Dataset

This dataset is built specifically for smart TV gesture control systems and contains about 500 independently collected short video samples. Each video clip lasts 2 to 3 seconds, fully recording the dynamic process from the start of the gesture to the full display. These gestures include thumbs up, thumbs up, swipe left, swipe right, and stop, and serve as separate training samples for gesture recognition models. The samples were collaboratively completed by participants of different ages (18-65 years old), genders, and skin colors, covering a variety of interactive postures such as standing and sitting to capture the differences in operating habits that may occur in real users. The original video resolution is 640×480 pixels, simulating the imaging characteristics of mainstream home webcams to ensure the consistency of training data with real application scenarios.