EXTD: Extremely Tiny Face Detector via Iterative Filter Reuse

Abstract

In this paper, we propose a new multi-scale face detector having an extremely tiny number of parameters (EXTD),less than 0.1 million, as well as achieving comparable performance to deep heavy detectors. While existing multi-scale face detectors extract feature maps with different scales from a single backbone network, our method generates the feature maps by iteratively reusing a shared lightweight and shallow backbone network. This iterative sharing of the backbone network significantly reduces the number of parameters, and also provides the abstract image semantics captured from the higher stage of the network layers to the lower-level feature map. The proposed idea is employed by various model architectures and evaluated by extensive experiments. From the experiments from WIDER FACE dataset, we show that the proposed face detector can handle faces with various scale and conditions, and achieved comparable performance to the more massive face detectors that few hundreds and tens times heavier in model size and floating point operations.

References in corpus (10)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• YOLO9000: Better, Faster, Stronger
• Light-Head R-CNN: In Defense of Two-Stage Object Detector
• Face Attention Network: An Effective Face Detector for the Occluded Faces
• Face Detection through Scale-Friendly Deep Convolutional Networks
• Detecting Faces Using Region-based Fully Convolutional Networks
• DFANet: Deep Feature Aggregation for Real-Time Semantic Segmentation
• Depthwise Convolution is All You Need for Learning Multiple Visual Domains
• Improved Selective Refinement Network for Face Detection
• ShaResNet: reducing residual network parameter number by sharing weights

Cited by in corpus (2)

• Face Detection with Feature Pyramids and Landmarks
• SINet: Extreme Lightweight Portrait Segmentation Networks with Spatial Squeeze Modules and Information Blocking Decoder