-
Numpy
- Numpy cheatsheet
-
Pandas
- Dataframe shuffle
- DataFrame to list of dictionary
- DataFrame filter to select data
-
Matplotlib
- subplots
- pie plot
-
Magic Python
- Model
- Model Backbone
- Reset Family(NFnet、Resxnet、Resnset...)
- Repvgg
- Gan
- Enhance Model
- Attention Layer
- Spatial Pyramid Pooling (SPP Layer)
- Feature fusion / Module fusion
- Model Backbone
- Training Trick
- Cross-validation
- Ensemble Learning
- Test Time Augmentation
- Noise Label
- Training in Noise Labels
- Label smoothing
- Pytorch
- Seed setting
- Modify pre-trained model
- Fine-tune model
- Improve speed of pytorch training
- Distributed training
- Learning scheduler
- Activation
- Computer Vision
- Grad-CAM
- Kaggle Trick
- albumentations
- Recommend Website
- Exploratory Data Analysis(EDA)
- Text to Value
- One hot encoding
- Normalization
- Zero-mean normalization
- Instance Normalization
- Image Convert
- Gray Level Co-occurrence Matrix(GLCM) 灰度共生矩陣
- Image Decompose
- Image Augmentation
- Albumentations Library
- Image Contrast Enhancement
- Mixup/Cutout/CutMix Training Data
- Dataset Format
- VOC to COCO
- COCO DataLoader
- Data Vision
- T-SNE
- Model
- XgBoost
- SVM
Data Science大全,從Git、SQL、Excel、Algorithms...都有
找不到資料或是lib都來這裡
很多競賽跟資料集,可以去看別人的notebook來學習,很多人會在上面寫教學
版主文章質量很高,我大部分資源都從這裡取得
找model、找Task,來這就對了,論文 + github code,最近還釋出3000+ Dataset,任務分類詳細,可以找到最新的模型架構跟基準
跟 paper with code 差不多,可以看到所有open source model 的 benchmarks,所以可以在上面找到目前主流的模型
這是一個知乎用戶,常常會發佈關於Computer vision的最新研究
你可以在任何github repositories的網址上,把github.com
改成 github1s.com
,就可以用vscode editor查看
看論文可以找 《XX 論文筆記》
吐血推薦,B站最強學習資源彙總(數據科學,機器學習,python)
-
NCCU自學課程-成為python數據分析達人的第一課
- Tensorflow 教學 +Code
- Tensorflow2.0 最強教學!
- Tensorflow2.0 開源教學電子書
- Tensorflow2.0 《30天吃掉那隻TensorFlow2.0 》最強教學!
- TI邦幫忙-Towards Tensorflow 2.0 30Days
- Deepfake Tutorial Website
- deepfake實戰教學