Skip to main content

Table 3 Summary of our dataset used for evaluation

From: First-person reading activity recognition by deep learning with synthetically generated images

Recording devices HX-A5002 Looxcie23
Image size 960 × 540 480 × 320
Number of videos 3 videos 17 videos
Sampling rate 5–6 fps
Number of images (per video) 100–700 per a video