FaceRec
簡單易懂，高精準度的人臉辨識技術封裝
Papers
深度學習人臉辨識技術
-
基礎: 機器學習/深度學習/圖形處理器技術
-
"DeepFace: Closing the Gap to Human-Level Performance in Face Verification"
最早的深度學習人臉辨識, 已有 metric learning 的觀念 (使用 siamese network)
但, 無權值共享的 CNN 帶來過多的參數, 3D alignment 也顯得過度複雜
- "Deep Face Recognition"
*http://cis.csuohio.edu/~sschung/CIS660/DeepFaceRecognition_parkhi15.pdf
著名的 VGG Face, 整套流程包含 face dataset 的建立
- "FaceNet: A Unified Embedding for Face Recognition and Clustering"
用 triplet loss 產生 128 維的 FaceNet embeddings (此向量空間內的距離代表人臉的相似程度), LFW 準確度超過 99%
網路結構:
- (A) "Very deep convolutional networks for large-scale image recognition. In International Conference on Learning Representations"
*https://arxiv.org/pdf/1409.1556/
經典的 VGG Network, 包含 VGG16, VGG19
- "Going Deeper With Convolutions"
http://openaccess.thecvf.com/content_cvpr_2015/papers/Szegedy_Going_Deeper_With_2015_CVPR_paper.pdf
GoogLeNet, 使用 3x3, 1x1 convolution 構成 inception 網路模組
- "Deep residual learning for image recognition"
http://openaccess.thecvf.com/content_cvpr_2016/papers/He_Deep_Residual_Learning_CVPR_2016_paper.pdf
residual network, 解決梯度消失問題, 讓訓練 100 (甚至1000) 層以上的深度學習變得容易
- "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications"
https://arxiv.org/abs/1704.04861
mobile net, 小而快的網路， 但犧牲準確度，
A. "Labeled Faces in the Wild: A Database for Studying Face Recognition in Unconstrained Environments"
http://cs.brown.edu/courses/cs143/2011/proj4/papers/lfw.pdf
*著名的 lfw 人臉辨識準確率測試資料集
Results
99%
*https://github.com/BIG-CHENG/FaceRec/blob/master/fr_lfw_prec_recall_all.png
1 maintainer
Metadata
1 maintainer
Release files for frapi 0.1.6
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| frapi-0.1.6.tar.gz | 7.3 kB | Without provenance | Details |
Release files / frapi-0.1.6.tar.gz
| Download URL | frapi-0.1.6.tar.gz |
|---|---|
| Size | 7.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
d32906bec3764d9fdcdef41d898d30aa90d1c1b890be19663d36f67ca298c1a8
|
|
BLAKE2b-256 checksum How to use checksums |
5dd3f928bb7af24949d450c916b575d079773ea1fa2124e7b1e6cac344295bef
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/1.13.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/40.8.0 requests-toolbelt/0.9.1 tqdm/4.31.1 CPython/3.6.7
|
Security for frapi 0.1.6
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.
