Sample and Computation Redistribution for Efficient Face Detection

Overview

Introduction

SCRFD is an efficient high accuracy face detection approach which initially described in Arxiv.

prcurve

Performance

Precision, flops and infer time are all evaluated on VGA resolution.

ResNet family

Method Backbone Easy Medium Hard #Params(M) #Flops(G) Infer(ms)
DSFD (CVPR19) ResNet152 94.29 91.47 71.39 120.06 259.55 55.6
RetinaFace (CVPR20) ResNet50 94.92 91.90 64.17 29.50 37.59 21.7
HAMBox (CVPR20) ResNet50 95.27 93.76 76.75 30.24 43.28 25.9
TinaFace (Arxiv20) ResNet50 95.61 94.25 81.43 37.98 172.95 38.9
- - - - - - - -
ResNet-34GF ResNet50 95.64 94.22 84.02 24.81 34.16 11.8
SCRFD-34GF Bottleneck Res 96.06 94.92 85.29 9.80 34.13 11.7
ResNet-10GF ResNet34x0.5 94.69 92.90 80.42 6.85 10.18 6.3
SCRFD-10GF Basic Res 95.16 93.87 83.05 3.86 9.98 4.9
ResNet-2.5GF ResNet34x0.25 93.21 91.11 74.47 1.62 2.57 5.4
SCRFD-2.5GF Basic Res 93.78 92.16 77.87 0.67 2.53 4.2

Mobile family

Method Backbone Easy Medium Hard #Params(M) #Flops(G) Infer(ms)
RetinaFace (CVPR20) MobileNet0.25 87.78 81.16 47.32 0.44 0.802 7.9
FaceBoxes (IJCB17) - 76.17 57.17 24.18 1.01 0.275 2.5
- - - - - - - -
MobileNet-0.5GF MobileNetx0.25 90.38 87.05 66.68 0.37 0.507 3.7
SCRFD-0.5GF Depth-wise Conv 90.57 88.12 68.51 0.57 0.508 3.6

X64 CPU Performance of SCRFD-0.5GF:

Test-Input-Size CPU Single-Thread Easy Medium Hard
Original-Size(scale1.0) - 90.91 89.49 82.03
640x480 28.3ms 90.57 88.12 68.51
320x240 11.4ms - - -

precision and infer time are evaluated on AMD Ryzen 9 3950X, using the simple PyTorch CPU inference by setting OMP_NUM_THREADS=1 (no mkldnn).

Installation

Please refer to mmdetection for installation.

  1. Install mmcv. (mmcv-full==1.2.6 and 1.3.3 was tested)
  2. Install build requirements and then install mmdet.
    pip install -r requirements/build.txt
    pip install -v -e .  # or "python setup.py develop"
    

Pretrained-Models

Name Easy Medium Hard FLOPs Params(M) Infer(ms) Link
SCRFD_500M 90.57 88.12 68.51 500M 0.57 3.6 download
SCRFD_1G 92.38 90.57 74.80 1G 0.64 4.1 download
SCRFD_2.5G 93.78 92.16 77.87 2.5G 0.67 4.2 download
SCRFD_10G 95.16 93.87 83.05 10G 3.86 4.9 download
SCRFD_34G 96.06 94.92 85.29 34G 9.80 11.7 download
SCRFD_500M_KPS 90.97 88.44 69.49 500M 0.57 3.6 download
SCRFD_2.5G_KPS 93.80 92.02 77.13 2.5G 0.82 4.3 download
SCRFD_10G_KPS 95.40 94.01 82.80 10G 4.23 5.0 download

mAP, FLOPs and inference latency are all evaluated on VGA resolution. _KPS means the model includes 5 keypoints prediction.

Convert to ONNX

Please refer to tools/scrfd2onnx.py

Generated onnx model can accept dynamic input as default.

You can also set specific input shape by pass --shape 640 640, then output onnx model can be optimized by onnx-simplifier.

Inference

Put your input images or videos in ./input directory. The output will be saved in ./output directory. In root directory of project, run the following command for image:

python inference_image.py --input "./input/test.jpg"

and for video:

python inference_video.py --input "./input/obama.mp4"

Use -sh for show results during code running or not

Note that you can pass some other arguments. Take a look at inference_video.py file.

Owner
Sajjad Aemmi
AI MSc Student at Ferdowsi University of Mashhad - Teacher - Machine Learning Engineer - WebDeveloper - Graphist
Sajjad Aemmi
VQGAN+CLIP Colab Notebook with user-friendly interface.

VQGAN+CLIP and other image generation system VQGAN+CLIP Colab Notebook with user-friendly interface. Latest Notebook: Mse regulized zquantize Notebook

Justin John 227 Jan 05, 2023
This repo contains the official code of our work SAM-SLR which won the CVPR 2021 Challenge on Large Scale Signer Independent Isolated Sign Language Recognition.

Skeleton Aware Multi-modal Sign Language Recognition By Songyao Jiang, Bin Sun, Lichen Wang, Yue Bai, Kunpeng Li and Yun Fu. Smile Lab @ Northeastern

Isen (Songyao Jiang) 128 Dec 08, 2022
Image reconstruction done with untrained neural networks.

PyTorch Deep Image Prior An implementation of image reconstruction methods from Deep Image Prior (Ulyanov et al., 2017) in PyTorch. The point of the p

Atiyo Ghosh 192 Nov 30, 2022
This is the official PyTorch implementation of our paper: "Artistic Style Transfer with Internal-external Learning and Contrastive Learning".

Artistic Style Transfer with Internal-external Learning and Contrastive Learning This is the official PyTorch implementation of our paper: "Artistic S

51 Dec 20, 2022
Optimized code based on M2 for faster image captioning training

Transformer Captioning This repository contains the code for Transformer-based image captioning. Based on meshed-memory-transformer, we further optimi

lyricpoem 16 Dec 16, 2022
Meta-Learning Sparse Implicit Neural Representations (NeurIPS 2021)

Meta-SparseINR Official PyTorch implementation of "Meta-learning Sparse Implicit Neural Representations" (NeurIPS 2021) by Jaeho Lee*, Jihoon Tack*, N

Jaeho Lee 41 Nov 10, 2022
Liecasadi - liecasadi implements Lie groups operation written in CasADi

liecasadi liecasadi implements Lie groups operation written in CasADi, mainly di

Artificial and Mechanical Intelligence 14 Nov 05, 2022
multimodal transformer

This repo holds the code to perform experiments with the multimodal autoregressive probabilistic model Transflower. Overview of the repo It is structu

Guillermo Valle 68 Dec 13, 2022
Dilated Convolution for Semantic Image Segmentation

Multi-Scale Context Aggregation by Dilated Convolutions Introduction Properties of dilated convolution are discussed in our ICLR 2016 conference paper

Fisher Yu 764 Dec 26, 2022
Distance correlation and related E-statistics in Python

dcor dcor: distance correlation and related E-statistics in Python. E-statistics are functions of distances between statistical observations in metric

Carlos Ramos CarreƱo 108 Dec 27, 2022
SafePicking: Learning Safe Object Extraction via Object-Level Mapping, ICRA 2022

SafePicking Learning Safe Object Extraction via Object-Level Mapping Kentaro Wad

Kentaro Wada 49 Oct 24, 2022
In this project we combine techniques from neural voice cloning and musical instrument synthesis to achieve good results from as little as 16 seconds of target data.

Neural Instrument Cloning In this project we combine techniques from neural voice cloning and musical instrument synthesis to achieve good results fro

Erland 127 Dec 23, 2022
Deep Unsupervised 3D SfM Face Reconstruction Based on Massive Landmark Bundle Adjustment.

(ACMMM 2021 Oral) SfM Face Reconstruction Based on Massive Landmark Bundle Adjustment This repository shows two tasks: Face landmark detection and Fac

BoomStar 51 Dec 13, 2022
This code is 3d-CNN model that can predict environmental value

Predict-environmental-value-3dCNN This code is 3d-CNN model that can predict environmental value. Firstly, I built a model that can create a lot of bu

1 Jan 06, 2022
AISTATS 2019: Confidence-based Graph Convolutional Networks for Semi-Supervised Learning

Confidence-based Graph Convolutional Networks for Semi-Supervised Learning Source code for AISTATS 2019 paper: Confidence-based Graph Convolutional Ne

MALL Lab (IISc) 56 Dec 03, 2022
Implementation of Multistream Transformers in Pytorch

Multistream Transformers Implementation of Multistream Transformers in Pytorch. This repository deviates slightly from the paper, where instead of usi

Phil Wang 47 Jul 26, 2022
NVIDIA Deep Learning Examples for Tensor Cores

NVIDIA Deep Learning Examples for Tensor Cores Introduction This repository provides State-of-the-Art Deep Learning examples that are easy to train an

NVIDIA Corporation 10k Dec 31, 2022
Split your patch similarly to `git add -p` but supporting multiple buckets

split-patch.py This is git add -p on steroids for patches. Given a my.patch you can run ./split-patch.py my.patch You can choose in which bucket to p

102 Oct 06, 2022
LegoDNN: a block-grained scaling tool for mobile vision systems

Table of contents 1 Introduction 1.1 Major features 1.2 Architecture 2 Code and Installation 2.1 Code 2.2 Installation 3 Repository of DNNs in vision

41 Dec 24, 2022
PyTorch code for our paper "Image Super-Resolution with Non-Local Sparse Attention" (CVPR2021).

Image Super-Resolution with Non-Local Sparse Attention This repository is for NLSN introduced in the following paper "Image Super-Resolution with Non-

143 Dec 28, 2022