The scenario
Augment a raw fall-detection video frame dataset by applying synthetic occlusion and lighting variations, then validate and store the augmented samples.
What is in this drawing
Read the decisions behind it.
01
Apply occlusion augmentation
02
Apply lighting variation
03
Quality check passed
Use when building a training set for vision-based fall detection that needs robustness to occlusions and lighting changes.
More like this