Crossmodality Person Reidentification Based on Global and Local Alignment

Table 2

Overview of our dataset. The red part represents IR images converted from RGB images taken by 1st, 2nd, 4th, and 5th RGB cameras in SYSU-MM01 dataset through the AlignGAN model, while the black part represents all images in SYSU-MM01 dataset, including RGB images taken by 1st, 2nd, 4th, and 5th RGB cameras and IR images taken by 3rd and 6th IR cameras.

Cam	Location	(in/out) door	Lighting	ID#	RGB#/ID	IR#/ID
1	Room 1	Indoor	Bright	259	400+	—
2	Room 2	Indoor	Bright	259	400+	—
4	Gate	Outdoor	Bright	493	20	—
5	Garden	Outdoor	Bright	502	20	—
1	Room 1	Indoor	Dark	259	—	400+
2	Room 2	Indoor	Dark	259	—	400+
4	Gate	Outdoor	Dark	493	—	20
5	Garden	Outdoor	Dark	502	—	20
3	Room 2	Indoor	Dark	486	—	20
6	Passage	Outdoor	Dark	299	—	20