 PMPF, Point Cloud Multiple Pixel Fusion, is a novel, efficient, and effective multi-sensor fusion detection framework for 3D object detection. It is designed to overcome the drawbacks of existing LiDAR-only detectors by projecting the Point Cloud data onto the image plane and applying it to LiDAR-only detectors with autoencoders. This allows for a decoupled, plug-and-play approach which results in improved performance compared to other LiDAR-only detectors. Experimental results on the Kitty 3D object detection benchmark demonstrate that PMPF outperforms all other LiDAR-only detectors, as well as several two-stage detectors. This article was authored by Yan Zhang, Kang Lu, Hong Bao, and others.