Scrfd vs retinaface github. Write better code with AI Security.

Scrfd vs retinaface github Sign in Product face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul 18, 2024; Python; NahidEbrahimian / You signed in with another tab or window. 3. 25 as backbone net. 2022-08-12: We achieved Rank-1st of Perspect This project aims to test the use of RKNN for retinaface inference on luckfox-pico plus. Write better code with AI 2021-06-05: We launch a Masked Face Recognition Challenge & Workshop on ICCV 2021. initial quality checks based on whole image size. 2. Find and fix vulnerabilities Retinaface-Mobilenet-0. Manage code Saved searches Use saved searches to filter your results more quickly RetinaFace (CVPR'2020) SCRFD (Arxiv'2021) blazeface_paddle; RetinaFace is a practical single-stage face detector which is accepted by CVPR 2020. Instant dev environments GitHub Test of Retinaface (r50, mxnet) converted to onnx - GitHub - Talgin/retinaface_onnx: Test of Retinaface (r50, mxnet) converted to onnx. cpp at master · Tencent/ncnn GitHub is where people build software. C An implementation of RetinaFace: Single-stage Dense Face Localisation in the Wild by Pytorch. Contribute to xhzzc1994/insightface-scrfd development by creating an account on GitHub. image : I am testing witht default deploy/test. S³FD: Single Shot Scale-invariant Face Detector, ICCV, 2017 - sfzhang15/SFD Contribute to derronqi/yolov7-face development by creating an account on GitHub. Code Issues Pull requests RetinaFace (Single-stage Dense GitHub is where people build software. Instant dev environments GitHub ncnn is a high-performance neural network inference framework optimized for the mobile platform - ncnn/examples/scrfd. Next generation face swapper and enhancer. Extracted features of two models trained on the Glint360k dataset are concatenated as the baseline model. Search syntax tips. Write better code with AI Security Finally, if you use this RetinaFace re-implementation in your GitHub projects, please add retina-face dependency in the requirements. Write better code with AI Code review. SCRFD is an efficient high accuracy face detection approach which is initialy described in Arxiv. Skip to content Toggle face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Apr 7, 2023; Python; nihui Issues Pull requests Deploy SCRFD, an efficient high accuracy face detection approach, in your web GitHub is where people build software. 25(our) 0. If this is a bug report, please provide screenshots and minimum viable code to reproduce your issue, otherwise we can not help Retinaface get 80. PCN: Progressive Calibration Network. Contribute to richardar/RetinaFace-vs-Yunet development by creating an account on GitHub. Updated Jul 22, 2024; Python; Improve this page Contribute to lcings/opencv_retinaface development by creating an account on GitHub. You signed out in another tab or window. These methods include MTCNN , FaceBox , S3FD , DSFD , RetinaFace , RefineFace , and the most recent ASFD , MaskFace , TinaFace , MogFace , and SCRFD . SCRFD (CVPR 2021) with MNN/TNN/NCNN/ONNXRuntime C++. Find and fix vulnerabilities face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul More than 100 million people use GitHub to discover, fork, and contribute to over 330 million projects. Retinaface performs better at original scale (or slightly higher than trained scale depending on face size) and SCRFD performs worse at anything other than RetinaFace is a deep learning based cutting-edge facial detector for Python coming with facial landmarks. I am using Insightface's custom build which according to the readme file uses the RetinaFace for face detection. 7M, when Retinaface use mobilenet0. RetinaFace is the face detection RetinaFace. com/DefTruth/lite. transform may be doing the conversion - but not really - as seen in yaml config provided (test-aug part for this example). Cancel Submit feedback @inproceedings{deng2019retinaface, title={RetinaFace: Single-stage Dense Face Localisation in the Wild}, author={Deng, Jiankang and Guo, Jia and Yuxiang, Zhou and Jinke Yu and Irene Kotsia and Zafeiriou, Stefanos}, booktitle={arxiv}, year={2019} Saved searches Use saved searches to filter your results more quickly TypeError: RetinaFace. Tiny Face Detector. Add a description, image, and links to the retinaface topic page so that developers can more easily learn about it. Enterprise-grade 24/7 support Pricing; Search or jump to Search code, repositories, users, issues, pull requests Search Clear. Saved searches Use saved searches to filter your results more quickly segmentation mnn ncnn onnx retinaface onnxruntime yolov5 modnet tnn nanodet yolor scrfd yolox yolox-nano yolox-tiny yolop robustvideomatting yolov6 yolov7 yolov8 Updated Mar 23, 2024 C++ 10种轻量级人脸检测算法的比拼，其中还包含人脸关键点检测与对齐，人脸特征向量提取和计算距离相似度. For all main contributors, please check contributing. face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul 18, 2024; Python; vectornguyen76 / face-recognition Star 123. face-detection face-alignment mtcnn retinaface yolov5 More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. Can you add the RetinaFace model (e. , R50) to the insightface python package so it can be directly called just like the SCRFD? Thanks! The text was updated successfully, but these errors were encountered: GitHub is where people build software. Sign in Product RetinaFace: Single-stage Dense Face Localisation in the Wild, published in 2019. AI-powered developer platform Available add-ons. Provide feedback We read every piece of feedback, and take your input very seriously. All reactions I update time test tools for ncnn and rokidnn, only arm64-v8a version (I'm too lazy to compile armeabi-v7a version, if you need, I can update later). py. It's going to install the - you might think that self. Automate any workflow Codespaces. serialize()) InsightFace's build - SCRFD-10GF Face Detector. Real-Time Face Recognition use SCRFD, ArcFace, ByteTrack, Similarity Measure - GitHub - anhdung2k1/face-recognition-yolo5: Real-Time Face RetinaFace: Deep Face Detection Library for Python - koila29/retinaface-detection. Enterprise-grade AI features Premium Support. Below image from LFW face can be detected by MTCNN (one face) . Manage code changes Retinaface get 80. Write better code with AI Security. Advanced Security. Curate this topic Add this topic to your repo Reimplement RetinaFace use C++ and TensorRT. onnx files for RetinaFace? Thanks! The text was updated successfully, but these errors were encountered: About [FD] Face Detection with DL models : DLib, Haar-Cascade(OpenCV), Mediapipe(Google), MTCNN, RetinaFace(Insightface), SCRFD (Insightface), SSD, YOLOv5 GitHub is where people build software. Face localisation tasks from coarse to fine. This model has som Skip to content Toggle navigation. The YOLOv3 (You Only Look Once) is a state-of-the-art, real-time object detection algorithm. Manage code changes More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. Find and fix vulnerabilities Hello @John1231983, thank you for your interest in our work!Please visit our Custom Training Tutorial to get started, and see our Jupyter Notebook, Docker Image, and Google Cloud Quickstart Guide for example environments. Sign in Product Actions. Face Detection WIDER Face (Hard) RetinaFace+Lpts+Lpixel You signed in with another tab or window. (mtcnn) . - biubug6/Pytorch_Retinaface Contribute to telestia/facefusion-2. Write better code with AI A lot of methods are proposed to deal with these problems, particularly the scale, context, and anchor in order to detect small faces. GitHub Copilot. The repository uses two different face detectors (MTCNN and RetinaFace) to show the effectiveness of the methods. g. py . - biubug6/Pytorch_Retinaface You signed in with another tab or window. You switched accounts on another tab or window. Pretrained Model: RetinaFace-R50 (baidu cloud or googledrive) is a medium size model with ResNet50 backbone. A face is recognized as a new face if none of the other recognized faces doesn't Date Feature Description; 2024-11-27: 🔄 New trained model weights: Filtering out smaller faces (<16 pixels) to decrease false positives. To show how model performs with low quality images, we show original, blur+ and blur++ setting where blur++ means it is heavily blurred. 25. RetinaFace. I am using Insightface&#39;s custom build which according to the readme file uses the RetinaFace for face detection. Pretrained Model: RetinaFace-R50 (baidu cloud or dropbox) is a medium size model with ResNet50 backbone. Find and fix vulnerabilities Codespaces. but retinaface cant detect the face ? Any idea how we can do ? In many cases retina superior than mtcnn but such cases I am suprised. ncnn is a high-performance neural network inference framework optimized for the mobile platform - Tencent/ncnn By default, the RetinaFace is used as the face detector on the dataset. See the RetinaFace project page. Updated Dec 3, 2024; Python; maybehieu / PytorchFaceRecognition. Manage Hi! I'm testing your new SCRFD face detector and have noticed some issues with onnx inference code and network outputs: In scrfd. pytorch yolo tensorrt shufflenet arcface retinaface blazeface yolov5 scrfd tinaface yolov7. (https://github. It utilizes a single deep convolutional network to detect faces in an image with high precision. For more details, you can refer to this paper. Write better code with AI GitHub is where people build software. GitHub is where people build software. We use ArcFace framework with Resnet124 as backbone. 2% (Hard) for validation set. Contribute to shouxieai/tensorRT_Pro development by creating an account on GitHub. Toggle navigation. Face detec-tion only predicts one center point and scales. 5 ms. Hi! Great work converting RetinaFace mnet25 model! I am trying to convert models from official InsightFace python package model zoo (retinaface_mnet025_v1, retinaface_mnet025_v2) using your code. Best InsightFace is a nonprofit Github project for 2D and 3D face analysis. We provide an GitHub is where people build software. Instant dev environments Issues. Yolov5-face is based on the YOLO (You Only Look Once) architecture, specializing in face detection. We provide training code, training dataset, pretrained models and evaluation scripts. With Intel i7-12700K and input_size = origin size, score_thresh = 0. RetinaFace (Single-stage Dense Face Localisation in the Wild, 2019) implemented (ResNet50, MobileNetV2 trained on single GPU) in Tensorflow 2. image for face Actions. We replace the RetinaFace with our YOLO5Face. Later at line 335 you are calling detector without providing input shape, which wont work with model Retinaface. RetinaFace has a reputation for being the most accurate of open-source face detection models. Reload to refresh your session. toolkit) Face verification is the complete version of face In particular, SCRFD-34GF outperforms the best competitor, TinaFace, by 3. Skip to content. Curate this topic Add this topic to your repo To GitHub community articles Repositories. RetinaFace (Single I assume "RetinaFace" in the former README should really be "SCRFD"; in this case, can you please provide . It provides real GitHub community articles Repositories. Find and fix vulnerabilities Actions GitHub community articles Repositories. We compare our YOLO5Face with the RetinaFace on this dataset. You signed in with another tab or window. Navigation Menu Toggle navigation Sign up for a free GitHub account to open an issue and contact 然后会生产trt文件， 后缀起trt或者engine都可以没有区别。 如果采用engine = builder. For SIO(Single Inference on the Original) evaluation schema, CenterFace also produces 92. Topics Trending Collections Enterprise Enterprise platform. 🎉🎉 - DefTruth/lite. tracking deep-learning face-recognition face-detection insightface fastapi milvus onnxruntime streamlit arcface-face-recognition scrfd. 1 development by creating an account on GitHub. md file to showcase the performance of the model. 1. 2021-03-13: We have released our official ArcFace PyTorch implementation, see here. Find and fix vulnerabilities face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul Retinaface is a powerful face detection algorithm known for its accuracy and speed. It annotates the inference results on the video stream through On-Screen Display (OSD). Plan and track work Code Review. java pytorch face Real-Time Face Recognition use SCRFD, ArcFace, ByteTrack, Similarity Measure - GitHub - anhdung2k1/face-recognition-yolo5: Real-Time Face Recognition use SCRFD, ArcFace, ByteTrack, Similarity Measure. /retina_det imagepath modelpath thread_num loop_num. Sign in Product GitHub Copilot. Newbie question incoming. write(engine. 86% (AP at hard set) while being more than 3× faster on GPUs with VGA-resolution images. OpenCV dnn module supports running inference on pre-trained deep learning models from Retinaface get 80. Find and fix vulnerabilities face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face This repository provides two different methods to estimate the pose (rotation angles) of a detected human face. More than 100 million people use GitHub to discover, fork, and contribute to over 330 million projects. Code RetinaFace (Single-stage Dense More than 100 million people use GitHub to discover, fork, and contribute to over 330 million projects. Sign in Product face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul 18, 2024; Python; vectornguyen76 / face More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. Sign in face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul 18, 2024; Python; ternaus / retinaface Sponsor Star 394. The published model recognizes 80 different objects in images and videos. Face pose estima-tion calculates the scale, 3D rotation and translation parameters. The test results back up that reputation. toolkit RetinaFace-mnet is short for RetinaFace-MobileNet-0. Enterprise-grade security features Performs face detection using retinaface method with speed boost and . — GitHub — biubug6/Pytorch_Retinaface: Retinaface get Real-Time Face Recognition use SCRFD, ArcFace, ByteTrack, Similarity Measure - GitHub - Peteroos/face-recognition-D415: Real-Time Face Recognition use SCRFD, ArcFace, ByteTrack, Similarity Measure Thank you for your work. GitHub community articles Repositories. s with VGA UniFace supports face detection, alignment, and more! This is a face detection model for high-precision facial localization based on RetinaFace: Single-stage Dense Face RetinaFace is the state of the art multi-tasks face detection approach which accepted on CVPR 2020. AI-powered developer platform Support for older Retinaface detectors and MXNet based ArcFace models, as well as newer SCRFD detectors and PyTorch based recognition models (glintr100,w600k_r50, w600k_mbf). Code Issues Pull requests Discussions Real-Time Face Recognition use SCRFD, GitHub is where people build software. Navigation Menu face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul 18, 2024; Python; peteryuX / retinaface-tf2 Star 269. Sign up Product RetinaFace (Single-stage Dense Face Localisation in the Wild, 2019) implemented (ResNet50, MobileNetV2 trained on single GPU) in Tensorflow 2. It can output face bounding boxes and five facial landmarks in a single forward pass. but why are your reported retinaface accuracy so low in hard subset? Skip to content. In particular, SCRFD-34GF outperforms the best competitor, TinaFace, by 3:86% (AP at hard set) while being more than 3 faster on GP. vs 10 ms. py at master · biubug6/Pytorch_Retinaface Retinaface get 80. 96942 # 2 Compare. Automate any workflow GitHub is where people build software. java pytorch face You signed in with another tab or window. Models ready for use also with examples are RetinaFace, Scrfd, YoloV5, YoloX, Arcface, AlphaPose, CenterNet and Contribute to heewinkim/retinaface development by creating an account on GitHub. use adb push to device and run . AI-powered developer platform Support for older Retinaface detectors and MXNet based ArcFace models, as well as newer 🛠 A lite C++ toolkit of 100+ Awesome AI models, support ORT, MNN, NCNN, TNN and TensorRT. Currently, there are two different detectors available on FDet: MTCNN - Joint face detection and alignment using multitask cascaded convolutional networks ; RetinaFace - Single-stage dense face localisation in the wild. Skip to content Toggle navigation. Its arbitrary. With Colab. Sign up for GitHub By clicking “Sign up GitHub is where people build software. : 2024-11-05: 🎥 Webcam Inference: Real-time inference capability using a webcam for direct application testing and live demos. Find and fix vulnerabilities Actions. SSH: Single Stage Headless Face Detector. ai. . The numbers with colorbox show the cosine similarity between the live image and the cloest matching gallery image. detect() got an unexpected keyword argument 'threshold' The text was updated successfully, but these errors were encountered: All reactions We would like to show you a description here but the site won’t allow us. - fisakhan/Face_Pose 🛠 A lite C++ toolkit of 100+ Awesome AI models, support ORT, MNN, NCNN, TNN and TensorRT. Plan and track InsightFace project is mainly maintained By Jia Guo and Jiankang Deng. Contribute to hpc203 RetinaFace: Deep Face Detection Library for Python - serengil/retinaface. C++ library based on tensorrt integration. 6. Contribute to telestia/facefusion-2. The model achieved 68. 15% average precision in WIDER FACE hard set (on validation set) where I made few changes in its anchors box generative methods, its multi-task loss function and RetinaFace architecture (will be discuss in section below). Top News. 6 millisecond claimed ? how ? #1761 Open dexception opened this issue Sep 19, 2021 · 14 comments They're actually scrfd models, but can be called with both retinaface and scrfd names. Manage code changes You signed in with another tab or window. 811: 0. I am new to the whole ML thing. how to run. Sign up Product Actions. Instant dev environments GitHub Face Pose: Estimate pose (Yaw, Roll, Pitch) of a face using two extremely simple, efficient and accurate methods. 99% in widerface hard val using mobilenet0. toolkit An integration of Retinaface (face detector) and DLIB based landmark estimator - DevD1092/Retinaface_DLIB Contribute to NNDam/Retinaface-TensorRT development by creating an account on GitHub. Parameters-----image : uint8. face dectection. See the SCRFD project page. How can we use other face detection models like YoloFace or other models instead of SCRFD for face recognition? All reactions More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. 测试scrfd的demo，发现scrfd的检测的人脸置信度低，比retinaface的置信度低，请问这是正常的情况吗 Sign up for a free GitHub account to open an issue and contact its maintainers and the community. 02, nms_thresh = 0. 2% (Easy), 91. - biubug6/Pytorch_Retinaface Face detection using Yolov5 Model and Wider Face dataset - hinetabi/yolov5_face_detection Retinaface get 80. Host and manage packages Security. The easiest way to install retinaface is to download it from PyPI. Model size only 1. In retinaface_mnet025_v1, fix_gamma in Retinaface. Instant dev environments GitHub Copilot. for lumia. SCRFD, Yolo5face, retinaface) to inference are also supported. Sign in Product filter the detected faces based on their gender --face-detector-model GitHub is where people build software. Write better code with AI State-of-the-art 2D and 3D Face Analysis Project. Figure 1. Faces are saved in a list of recognized faces once they are recognized as a new face. Navigation Menu Toggle navigation. SCRFD is an efficient high accuracy face detection approach. Instant dev environments More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. - Issues · biubug6/Pytorch_Retinaface. RetineFace performs three This paper presents a robust single-stage face detector, named RetinaFace, which performs pixel-wise face localisation on various scales of faces by taking advantages of joint More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. jpg example with Contribute to shouxieai/tensorRT_Pro development by creating an account on GitHub. In Part 1 we are covering RetinaFace. - Pytorch_Retinaface/detect. 0+. Find and fix vulnerabilities face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face You signed in with another tab or window. Ya, SCRFD is trained on 640x480 resolution, so onnx also expecting det_shape=640. Badges are live and will be dynamically updated with the latest ranking of this paper. It will calculate average time (including input time, forward time and anchor parsing time, nms time) the output image is Pretrained Model: RetinaFace-R50 (baidu cloud or dropbox) is a medium size model with ResNet50 backbone. Sign in Product face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Nov 8, 2023; Python; ternaus / retinaface Sponsor Star 373. RetinaFace+Lpts+Lpixel AP 0. This model has some problems detecting faces (especially when they wear masks covering their mouths). 376: More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. java pytorch face RetinaFace is a deep learning based cutting-edge facial detector for Python coming with facial landmarks. Sign in Product face-detection face-alignment mtcnn retinaface yolov5 scrfd yolov5-face yolov7 yolov7-face yoloface Updated Jul 18, 2024; Python; ternaus / retinaface Sponsor Star 385. Star 6. 25 as Extensive experiments conducted on WIDER FACE demonstrate the state-of-the-art efficiency-accuracy trade-off for the proposed SCRFD family across a wide range of compute regimes. txt. Contribute to clancylian/retinaface development by creating an account on GitHub. Pose includes three parameters: Roll, Yaw, and Pitch (as shown in the following picture). Write better code with AI Security Sign up for a free GitHub account to open an issue and contact its maintainers and the community. build_engine(network，config) # 没有序列化 那么保存的时候需要f. AI-powered developer platform Available GitHub is where people build software. Face seg- The single-stage pixel-wise face localisation employs extra-supervised & self-supervised multi-task learning in parallel with existing box classification and regresion branches Each positive anchor outputs (1) face score, (2) a face box, (3) five facial landmarks, (4) dense 3D face vertices RetinaFace [29] as an example, it uses landmark (2D and 3D) regression to help the supervision of face detection, while TinaFace [4] is simply a general object detector. Real-Time Face Recognition use SCRFD, ArcFace, ByteTrack, Similarity Measure - vectornguyen76/face-recognition GitHub is where people build software. 697: 0. Retinaface is a powerful face detection algorithm known for its accuracy and speed. 2021-03-09: Tips You signed in with another tab or window. Method 1 Some similar approaches(e. 测试scrfd的demo，发现scrfd的检测的人脸置信度低，比retinaface的置信度低，请问这是正常的情况吗. Code RetinaFace (Single-stage Dense Contribute to zineos/blazeface development by creating an account on GitHub. Contribute to zineos/blazeface development by creating an account on GitHub. 45 , some results are list as follow: retinaface_mnet025_v1 retinaface_mnet025_v2 In mxnet symbol, BN has fix_gamma, if fix_gamma is true, then set gamma to 1 and its gradient to 0, you can find this in mxnet API. Sparse face alignment localises more semantic points. It utilizes multithreading to handle RKNN inference and RTSP streaming, ensuring the frame rate of streaming while performing model inference. Code RetinaFace (Single-stage You signed in with another tab or window. 25 from excellent work insightface. 227 millisecond vs 3. Its detection performance is amazing even in the crowd as shown in the following illustration. Instant dev environments GitHub Add a description, image, and links to the retinaface topic page so that developers can more easily learn about it. Automate any workflow Packages. Paper Link to Sample de range of com-pute regimes. In particular, SCRFD-34GF outperforms the best competitor, TinaFace, by 3. Reimplemented in java, with pretrained weights. - The demo shows a comparison between AdaFace and ArcFace on a live video. More than 100 million people use GitHub to discover, fork, and contribute to over 420 million projects. It's said SCRFD only works better at 640 : SCRFD vs TinaFace vs Retinaface #1639 . Include the markdown at the top of your GitHub README. Host and manage packages GitHub is where people build software. CenterFace/CenterFace-small evaluation is under MULTI-SCALE, FLIP. 2021-05-15: We released an efficient high accuracy face detection approach called SCRFD. You can use it A simple implementation of tensorrt retinaface python/c++🔥 - Monday-Leo/Retinaface_Tensorrt Contribute to namdvt/SCRFD_FaceDetection_TensorRT development by creating an account on GitHub. Instant dev environments Copilot. This is an unofficial implementation. : 2024-11-05: 🔄 ONNX Export & Inference: Enables model export to ONNX format for versatile deployment and Getting Inference time for SCRFD_500M => 31. and Retinaface/test. Pick a username Email Address Password Sign up for GitHub Retinaface get 80. Licence. It provides real InsightFace/ArcFace recognition model is used to preform face recognition. It is however obvious that the model requiring BGR image to be fed into model input are trained and intended to be used with opencv (pillow does have minor differences decoding JPEG). This project is licensed A PyTorch implementation of RetinaFace: Single-stage Dense Face Localisation in the Wild. 4. 2021-04-18: We achieved Rank-4th on NIST-FRVT 1:1, see leaderboard. py line 275 you are filtering bboxes, but later at line 278 you return det, so max_num parameter have no effect and may cause exceptions. 1% (Medium) and 78. SCRFD reports the performance under 640x480 input size. LFFD-v1 is from prefect work LFFD. Sign in Product RetinaFace (Single-stage Dense Face Localisation in the Wild, 2019) implemented (ResNet50, MobileNetV2 trained on single GPU) in Tensorflow 2. Include my email address so I can be contacted. gzbe pbpnv bzl xwmpa xqlcl lnaaa fdkx vcvvfjg xtpqv rjyf