kooose38's repositories
anime-face-classification
画像の分類とモデルの説明性を可視化します。
anime-segmentaion-project
Use the anime dataset to segment faces, hair and skins. In addition, since fine tuning is basically performed, evaluation by learning is not important. #segmentaion
pytools_table
Provides a class that defines effective tools for a dubbed dataset. Builds a model and generalizes preprocessing.
anime_dashboard
Sort the data obtained from api. Generate a dashboard with webui.
anomaly-detection-gan
Anomaly detection is performed from the image data using Gan. Here, we emphasize the flow of model construction and forward propagation. Therefore, I am not learning satisfactorily.
book_list
It will be a repository for taking notes on books.
Car-object-detective
物体検知とセグメンテーションを使った車窓からの予測実行する。画像をシーンで区分けして動画として出力する。
cpp_practice
c++の基本文法練習コード
disease-anaysis
Beginners competition in signate classification analysis pagas
Face-identification-system
自身の顔を登録。画像内から顔を検知し、登録された人物が存在するかを識別。複合モデルを用いた推論を構築する。
face-recognition-AI
Extract the face from the image file and make an inference using the similarity as an index.
gan-generate-image
Perform an image generation task using convolution. Use the most standard gan as the model.
image-caption-model
Generate sentences explained from images. However, the model is not trained, only the learning flow. Due to lack of gpu.
Image-composition
The segmentation model automatically detects the object and synthesizes the background image.
openpose-AI
Posture estimation is performed using various models. Output the image file from the execution command
petfinder
kaggle competition
pytools_nlp
By automating natural language processing, it can be transformed into a shape that can be handled by machine learning.
recommend-recipe-bot
We recommend recipes through line. It's a combination of a natural language processing engine and unsupervised tasks.
Sartorius-segmentaion-2021
kaggle competition segmentaion TASKS
semantic-segmentaion-pspnet
Extracts the features of an image by combining multiple convolution layers.
systematic-review-signate
signateコンペの訂正版。
tomorrows-corona-infections
Predict the number of infected people throughout Japan. Gay
youtube_api_dashboard
youtubeの視聴履歴から分析ダッシュボードを生成します。