Insects pose a significant threat to food security, destroying one-fifth of the global crop yield annually and impacting human and animal health by spreading diseases and damaging resources. Identifying insects traditionally relies on skilled taxonomists, but the diverse topography of regions like India makes this challenging. To address this, we propose a novel deep learning (DL) network combining the InceptionResNetV2 pre-trained model and Convolutional Block Attention Model (CBAM). The InceptionResNetV2 model is fine-tuned by unfreezing the last 5 blocks, allowing the model to learn from the given dataset and adjust its weights accordingly. The model achieves 84.46% classification accuracy. The Grad-CAM tool highlights key image regions for predictions, aiding farmers in distinguishing insects from crops and addressing crop-related issues. This integration of image processing, computer vision, and object detection enhances the efficiency of insect identification, benefiting agriculture and related industries.