What is group normalization and when is it preferred over batch normalization?

Group Normalization divides channels into groups and normalizes within each group, making it independent of batch size.

Batch Normalization depends on batch statistics, which becomes unstable for small batch sizes common in detection and segmentation. Group Normalization splits channels into groups and computes mean/variance within each group per sample. This makes it stable even when batch size is 1 and is widely used in dense vision tasks.