When should one train YOLOv5 on custom dataset or use standard COCO weights?

COCO dataset contains more than 91 stuff categories with 330K images (>200K labeled). Perhaps it is not the best way to identify people using YOLOv5 because your workshop has illumination changes during the daytime (natural light and lamps?). Thus, I suggest the following:

1. Use other pre-trained weights for people detection only, use this. I got them from this website, and they work pretty well.
2. Label people from recorded videos in different hours and check the recommendations below.

• Images per class. ≥ 1500 images per class recommended
• Instances per class. ≥ 10000 instances (labeled objects) per class recommended
• Image variety. Must be representative of deployed environment. For real-world use cases we recommend images from different times of day, different seasons, different weather, different lighting, different angles, different sources (scraped online, collected locally, different cameras) etc.
• Label consistency. All instances of all classes in all images must be labelled.
• Partial labelling will not work.
• Label accuracy. Labels must closely enclose each object. No space should exist between an object and it's bounding box. No objects should be missing a label.
• Background images. Background images are images with no objects that are added to a dataset to reduce False Positives (FP). We recommend about 0-10% background images to help reduce FPs (COCO has 1000 background images for reference, 1% of the total). No labels are required for background images.