Grid Loss: Detecting Occluded Faces
This addresses a specific challenge in face detection for applications like surveillance or real-time systems, though it is an incremental improvement over existing CNN methods.
The paper tackles the problem of detecting partially occluded faces in computer vision by proposing a novel grid loss layer for CNNs, which improves performance on benchmarks and achieves state-of-the-art results with no runtime cost.
Detection of partially occluded objects is a challenging computer vision problem. Standard Convolutional Neural Network (CNN) detectors fail if parts of the detection window are occluded, since not every sub-part of the window is discriminative on its own. To address this issue, we propose a novel loss layer for CNNs, named grid loss, which minimizes the error rate on sub-blocks of a convolution layer independently rather than over the whole feature map. This results in parts being more discriminative on their own, enabling the detector to recover if the detection window is partially occluded. By mapping our loss layer back to a regular fully connected layer, no additional computational cost is incurred at runtime compared to standard CNNs. We demonstrate our method for face detection on several public face detection benchmarks and show that our method outperforms regular CNNs, is suitable for realtime applications and achieves state-of-the-art performance.