基于Python的人脸识别自适应图片不同质量

上传者: Oliver9987 | 上传时间: 2023-03-22 16:40:13 | 文件大小: 35.81MB | 文件类型: ZIP
低质量人脸数据集中的识别具有挑战性,因为面部属性被遮挡和退化。基于边缘的损失函数的进步导致了嵌入空间中人脸的可辨别性增强。此外,先前的研究已经研究了自适应损失对错误分类(硬)示例赋予更多重要性的影响。在这项工作中,我们在损失函数中引入了自适应性的另一个方面,即图像质量。我们认为强调错误分类样本的策略应该根据它们的图像质量进行调整。具体来说,简单或困难样本的相对重要性应基于样本的图像质量。我们提出了一种新的损失函数,它根据图像质量强调不同难度的样本。我们的方法通过使用特征规范来近似图像质量,以自适应边缘函数的形式实现了这一点。大量实验表明,我们的方法 AdaFace 在四个数据集(IJB-B、IJB-C、IJB-S 和 TinyFace)上的人脸识别性能优于最先进的 (SoTA)。

文件下载

资源详情

[{"title":"( 89 个子文件 35.81MB ) 基于Python的人脸识别自适应图片不同质量","children":[{"title":"人脸识别","children":[{"title":"utils.py <span style='color:#111;'> 4.07KB </span>","children":null,"spread":false},{"title":"train_val.py <span style='color:#111;'> 10.15KB </span>","children":null,"spread":false},{"title":"人脸识别.txt <span style='color:#111;'> 950B </span>","children":null,"spread":false},{"title":"assets","children":[{"title":"adaface_demo5.gif <span style='color:#111;'> 16.44MB </span>","children":null,"spread":false},{"title":"main_figure.png <span style='color:#111;'> 180.28KB </span>","children":null,"spread":false},{"title":"video_thumbnail.png <span style='color:#111;'> 1.50MB </span>","children":null,"spread":false},{"title":"ms1mv2_train_subset_index.txt <span style='color:#111;'> 7.23MB </span>","children":null,"spread":false}],"spread":true},{"title":"main.py <span style='color:#111;'> 4.59KB </span>","children":null,"spread":false},{"title":"LICENSE <span style='color:#111;'> 1.04KB </span>","children":null,"spread":false},{"title":"validation_hq","children":[{"title":"eval_5valsets.sh <span style='color:#111;'> 1.50KB </span>","children":null,"spread":false}],"spread":true},{"title":"人脸识别.docx <span style='color:#111;'> 12.62KB </span>","children":null,"spread":false},{"title":"face_alignment","children":[{"title":"test_images","children":[{"title":"img3.jpeg <span style='color:#111;'> 184.59KB </span>","children":null,"spread":false},{"title":"img2.jpeg <span style='color:#111;'> 8.02KB </span>","children":null,"spread":false},{"title":"img1.jpeg <span style='color:#111;'> 4.48KB </span>","children":null,"spread":false}],"spread":true},{"title":"mtcnn_pytorch","children":[{"title":"src","children":[{"title":"__init__.py <span style='color:#111;'> 80B </span>","children":null,"spread":false},{"title":"matlab_cp2tform.py <span style='color:#111;'> 8.32KB </span>","children":null,"spread":false},{"title":"visualization_utils.py <span style='color:#111;'> 786B </span>","children":null,"spread":false},{"title":"first_stage.py <span style='color:#111;'> 3.11KB </span>","children":null,"spread":false},{"title":"weights","children":[{"title":"rnet.npy <span style='color:#111;'> 590.48KB </span>","children":null,"spread":false},{"title":"pnet.npy <span style='color:#111;'> 40.30KB </span>","children":null,"spread":false},{"title":"onet.npy <span style='color:#111;'> 2.24MB </span>","children":null,"spread":false}],"spread":false},{"title":"align_trans.py <span style='color:#111;'> 10.66KB </span>","children":null,"spread":false},{"title":"detector.py <span style='color:#111;'> 4.47KB </span>","children":null,"spread":false},{"title":"box_utils.py <span style='color:#111;'> 6.76KB </span>","children":null,"spread":false},{"title":"get_nets.py <span style='color:#111;'> 4.83KB </span>","children":null,"spread":false}],"spread":true},{"title":"LICENSE <span style='color:#111;'> 1.05KB </span>","children":null,"spread":false},{"title":"refine_faces.ipynb <span style='color:#111;'> 55.54KB </span>","children":null,"spread":false},{"title":"try_mtcnn_step_by_step.ipynb <span style='color:#111;'> 3.55MB </span>","children":null,"spread":false},{"title":"caffe_models","children":[{"title":"det1.prototxt <span style='color:#111;'> 2.30KB </span>","children":null,"spread":false},{"title":"det4.prototxt <span style='color:#111;'> 13.19KB </span>","children":null,"spread":false},{"title":"det3.prototxt <span style='color:#111;'> 3.84KB </span>","children":null,"spread":false},{"title":"det4.caffemodel <span style='color:#111;'> 3.62MB </span>","children":null,"spread":false},{"title":"det3.caffemodel <span style='color:#111;'> 1.49MB </span>","children":null,"spread":false},{"title":"det1.caffemodel <span style='color:#111;'> 27.50KB </span>","children":null,"spread":false},{"title":"det2.caffemodel <span style='color:#111;'> 398.35KB </span>","children":null,"spread":false},{"title":"det2.prototxt <span style='color:#111;'> 3.08KB </span>","children":null,"spread":false}],"spread":false},{"title":"extract_weights_from_caffe_models.py <span style='color:#111;'> 1.71KB </span>","children":null,"spread":false},{"title":"get_aligned_face_from_mtcnn.ipynb <span style='color:#111;'> 835.32KB </span>","children":null,"spread":false},{"title":".gitignore <span style='color:#111;'> 32B </span>","children":null,"spread":false},{"title":"images","children":[{"title":"example.png <span style='color:#111;'> 408.85KB </span>","children":null,"spread":false},{"title":"office4.jpg <span style='color:#111;'> 201.58KB </span>","children":null,"spread":false},{"title":"office5.jpg <span style='color:#111;'> 65.12KB </span>","children":null,"spread":false},{"title":"jf.jpg <span style='color:#111;'> 25.92KB </span>","children":null,"spread":false},{"title":"face0.jpg <span style='color:#111;'> 7.26KB </span>","children":null,"spread":false},{"title":"office3.jpg <span style='color:#111;'> 62.69KB </span>","children":null,"spread":false},{"title":"office1.jpg <span style='color:#111;'> 67.44KB </span>","children":null,"spread":false},{"title":"office2.jpg <span style='color:#111;'> 124.50KB </span>","children":null,"spread":false}],"spread":false},{"title":"README.md <span style='color:#111;'> 735B </span>","children":null,"spread":false},{"title":"test_on_images.ipynb <span style='color:#111;'> 3.28MB </span>","children":null,"spread":false}],"spread":false},{"title":"mtcnn.py <span style='color:#111;'> 6.70KB </span>","children":null,"spread":false},{"title":"align.py <span style='color:#111;'> 1.04KB </span>","children":null,"spread":false}],"spread":true},{"title":"convert.py <span style='color:#111;'> 3.58KB </span>","children":null,"spread":false},{"title":"dataset","children":[{"title":"record_dataset.py <span style='color:#111;'> 3.68KB </span>","children":null,"spread":false},{"title":"five_validation_dataset.py <span style='color:#111;'> 2.05KB </span>","children":null,"spread":false},{"title":"augmenter.py <span style='color:#111;'> 3.75KB </span>","children":null,"spread":false},{"title":"image_folder_dataset.py <span style='color:#111;'> 2.37KB </span>","children":null,"spread":false}],"spread":true},{"title":"inference.py <span style='color:#111;'> 1.35KB </span>","children":null,"spread":false},{"title":"head.py <span style='color:#111;'> 6.05KB </span>","children":null,"spread":false},{"title":"validation_mixed","children":[{"title":"insightface_ijb_helper","children":[{"title":"infer_helper.py <span style='color:#111;'> 4.86KB </span>","children":null,"spread":false},{"title":"eval_helper_all_proto.py <span style='color:#111;'> 13.96KB </span>","children":null,"spread":false},{"title":"recognition","children":[{"title":"__init__.py <span style='color:#111;'> 0B </span>","children":null,"spread":false},{"title":"embedding.py <span style='color:#111;'> 2.29KB </span>","children":null,"spread":false},{"title":"embedding_pytorch.py <span style='color:#111;'> 9.54KB </span>","children":null,"spread":false},{"title":"fp16_grad.py <span style='color:#111;'> 1.29KB </span>","children":null,"spread":false}],"spread":false},{"title":"dataloader.py <span style='color:#111;'> 3.97KB </span>","children":null,"spread":false},{"title":"eval_helper.py <span style='color:#111;'> 7.56KB </span>","children":null,"spread":false},{"title":"eval_helper_identification.py <span style='color:#111;'> 5.43KB </span>","children":null,"spread":false}],"spread":false},{"title":"validate_IJB_BC.py <span style='color:#111;'> 12.64KB </span>","children":null,"spread":false},{"title":"eval_ijb.sh <span style='color:#111;'> 1.42KB </span>","children":null,"spread":false},{"title":"README.md <span style='color:#111;'> 224B </span>","children":null,"spread":false}],"spread":false},{"title":"data.py <span style='color:#111;'> 8.90KB </span>","children":null,"spread":false},{"title":"requirements.txt <span style='color:#111;'> 81B </span>","children":null,"spread":false},{"title":"README_TRAIN.md <span style='color:#111;'> 3.91KB </span>","children":null,"spread":false},{"title":".gitignore <span style='color:#111;'> 103B </span>","children":null,"spread":false},{"title":"net.py <span style='color:#111;'> 12.49KB </span>","children":null,"spread":false},{"title":"README.md <span style='color:#111;'> 16.08KB </span>","children":null,"spread":false},{"title":"validation_lq","children":[{"title":"PFE","children":[{"title":"utils.py <span style='color:#111;'> 9.68KB </span>","children":null,"spread":false},{"title":"ijbs.py <span style='color:#111;'> 25.86KB </span>","children":null,"spread":false}],"spread":false},{"title":"eval_ijbs.sh <span style='color:#111;'> 259B </span>","children":null,"spread":false},{"title":"validate_tinyface.py <span style='color:#111;'> 6.39KB </span>","children":null,"spread":false},{"title":"evaluate_helper.py <span style='color:#111;'> 10.02KB </span>","children":null,"spread":false},{"title":"validate_IJB_S.py <span style='color:#111;'> 4.62KB </span>","children":null,"spread":false},{"title":"tinyface_helper.py <span style='color:#111;'> 9.32KB </span>","children":null,"spread":false},{"title":"README.md <span style='color:#111;'> 3.13KB </span>","children":null,"spread":false},{"title":"data_utils.py <span style='color:#111;'> 4.85KB </span>","children":null,"spread":false}],"spread":false},{"title":"evaluate_utils.py <span style='color:#111;'> 9.17KB </span>","children":null,"spread":false},{"title":"config.py <span style='color:#111;'> 3.93KB </span>","children":null,"spread":false},{"title":"scripts","children":[{"title":"run_ir50_ms1mv2.sh <span style='color:#111;'> 531B </span>","children":null,"spread":false},{"title":"run_ir101_ms1mv2.sh <span style='color:#111;'> 533B </span>","children":null,"spread":false}],"spread":false}],"spread":false}],"spread":true}]

评论信息

免责申明

【只为小站】的资源来自网友分享,仅供学习研究,请务必在下载后24小时内给予删除,不得用于其他任何用途,否则后果自负。基于互联网的特殊性,【只为小站】 无法对用户传输的作品、信息、内容的权属或合法性、合规性、真实性、科学性、完整权、有效性等进行实质审查;无论 【只为小站】 经营者是否已进行审查,用户均应自行承担因其传输的作品、信息、内容而可能或已经产生的侵权或权属纠纷等法律责任。
本站所有资源不代表本站的观点或立场,基于网友分享,根据中国法律《信息网络传播权保护条例》第二十二条之规定,若资源存在侵权或相关问题请联系本站客服人员,zhiweidada#qq.com,请把#换成@,本站将给予最大的支持与配合,做到及时反馈和处理。关于更多版权及免责申明参见 版权及免责申明