recognition gesture hand detection segmentation

The gap

Supporting evidence — 5 representative gaps

• Real-Time Hand Control for Interactive Presentation (2026) · doi

  Testing was conducted exclusively on static and dynamic hand gestures with only 20 gesture types from a controlled webcam setup; robustness evaluation is needed for real-time presentation scenarios with occlusions, fast hand movements, complex backgrounds, and varying distances from the camera.

  Keywords: hand gesture recognition real-time presentation robustness occlusion edge detection background segmentation
• Real-Time Hand Control for Interactive Presentation (2026) · doi

  The paper does not specify which trained model (KNN, HMM, or Markov Model) achieved the 96.5% accuracy; ablation studies comparing these three classifiers on the same hand gesture recognition task with skin segmentation and edge detection preprocessing are missing.

  Keywords: KNN HMM Markov Model hand gesture recognition classification skin detection edge detection
• Virtual Voice Assistant - for desktop (2026) · doi

  Existing methods for hand gesture recognition have limitations such as requiring a plain background with a good amount of light and stability.

  Keywords: existing hand gesture recognition limitations requiring plain background good amount light stability
• Virtual Voice Assistant - for desktop (2026) · doi

  We can reduce the error of recognition of gestures more than once for the same task by adding delay to the gesture.

  Keywords: reduce error recognition gestures once task adding delay gesture
• Vision and Voice Controlled Bionic Robotic Arm for Assistive and Research Applications (2026) · doi

  Gesture recognition using deep learning has not been incorporated into the current system design.

  Keywords: gesture recognition using deep learning incorporated current system design