Feng Sha

Title: Innovations of Feng Sha in Vehicle Detection Technology

Introduction

Feng Sha is a prominent inventor based in Zhejiang, China. He has made significant contributions to the field of deep learning, particularly in the area of vehicle detection. His innovative work addresses the challenges associated with detecting small-size vehicles in images.

Latest Patents

Feng Sha holds a patent for a "Small-size vehicle detection deep learning model based on feature fusion of multi-scale modules." This model effectively solves the problem of small-size vehicle image detection. It includes a Backbone network, a Neck layer, and a Head network. The Backbone network incorporates a C2f_DCNv3 module, which combines deformable convolution v3 (DCNv3) and a cross-stage feature fusion (C2f) module. Additionally, it features an SPPF_LSKA module that integrates a spatial pyramid pooling fast (SPPF) layer with a large separable kernel attention (LSKA) module. The Neck layer introduces a C2f_SCConv module, which combines spatial and channel reconstruction convolution (SCConv) with a C2f module. Finally, the Head network includes a multi-scale kernel detection (MSK_Detect) module.

Career Highlights

Feng Sha is currently associated with Nanhu Laboratory, where he continues to advance his research and development efforts. His work has garnered attention for its innovative approach to vehicle detection technology.

Collaborations

Feng collaborates with notable colleagues, including Yicheng Qiu and Li Niu, who contribute to his research initiatives.

Conclusion

Feng Sha's contributions to deep learning and vehicle detection technology highlight his innovative spirit and dedication to solving real-world problems. His patent reflects a significant advancement in the field, showcasing the potential of deep learning models in enhancing vehicle detection capabilities.