Microscopic target detection has always been a notorious challenge in object detection research, and we are proud to present our solution to this problem. Our proposed technique is based on YOLOv7, which is a foundational architecture that achieves both high accuracy and fast detection speeds. Our approach combines various techniques that include the integration of BiFormer attention mechanisms (Zhu et al, BiFormer: vision transformer with bi-level routing attention. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 10323–10333, 2023), the utilization of multiple convolutional layers for extracting richer feature information, and the application of non-maximum suppression principles and IoU calculations. These techniques enable us to focus on crucial feature information of small targets, which enhances detection accuracy while maintaining rapid detection speeds. To validate our proposed method, we conducted rigorous testing on a dedicated dataset. We assessed key metrics such as precision (P), recall (R), and mean average precision (mAP) to evaluate its performance. The results unequivocally demonstrate that our improved algorithm surpasses existing techniques and significantly enhances detection accuracy on the test dataset. We are confident that our proposed method will be a game changer in the field of microscopic target detection and will pave the way for further advancements in object detection research.