CAMAL: Context-Aware Multi-scale Attention framework for Lightweight Visual Place Recognition | Read Paper on Bytez