All Projects → zukakosan → street2shopTriplet

zukakosan / street2shopTriplet

Licence: other
Challenge the customer to shop task with tripletNet

Programming Languages

Jupyter Notebook
11667 projects

Projects that are alternatives of or similar to street2shopTriplet

Fashion-Clothing-Parsing
FCN, U-Net models implementation in TensorFlow for fashion clothing parsing
Stars: ✭ 29 (+107.14%)
Mutual labels:  fashion
facenet-pytorch-glint360k
A PyTorch implementation of the 'FaceNet' paper for training a facial recognition model with Triplet Loss using the glint360k dataset. A pre-trained model using Triplet Loss is available for download.
Stars: ✭ 186 (+1228.57%)
Mutual labels:  triplet-loss
cnn-for-image-retrieval
🌅The code of post "Image retrieval using MatconvNet and pre-trained imageNet"
Stars: ✭ 623 (+4350%)
Mutual labels:  image-retrieval
triplet
Re-implementation of tripletloss function in FaceNet
Stars: ✭ 94 (+571.43%)
Mutual labels:  triplet-loss
visual-compatibility
Context-Aware Visual Compatibility Prediction (https://arxiv.org/abs/1902.03646)
Stars: ✭ 92 (+557.14%)
Mutual labels:  fashion
Huawei DIGIX ImageRetri Top2
2020 DIGIX GLOBAL AI CHALLENGE - Digital Device Image Retrieval - Top2 WEARE队
Stars: ✭ 63 (+350%)
Mutual labels:  image-retrieval
CIRR
Official repository of ICCV 2021 - Image Retrieval on Real-life Images with Pre-trained Vision-and-Language Models
Stars: ✭ 65 (+364.29%)
Mutual labels:  image-retrieval
imageRetrieval
Image retrieval learning record
Stars: ✭ 31 (+121.43%)
Mutual labels:  image-retrieval
LabelRelaxation-CVPR21
Official PyTorch Implementation of Embedding Transfer with Label Relaxation for Improved Metric Learning, CVPR 2021
Stars: ✭ 37 (+164.29%)
Mutual labels:  image-retrieval
triplet-loss-pytorch
Highly efficient PyTorch version of the Semi-hard Triplet loss ⚡️
Stars: ✭ 79 (+464.29%)
Mutual labels:  triplet-loss
proxy-synthesis
Official PyTorch implementation of "Proxy Synthesis: Learning with Synthetic Classes for Deep Metric Learning" (AAAI 2021)
Stars: ✭ 30 (+114.29%)
Mutual labels:  image-retrieval
natural-language-joint-query-search
Search photos on Unsplash based on OpenAI's CLIP model, support search with joint image+text queries and attention visualization.
Stars: ✭ 143 (+921.43%)
Mutual labels:  image-retrieval
GPQ
Generalized Product Quantization Network For Semi-supervised Image Retrieval - CVPR 2020
Stars: ✭ 60 (+328.57%)
Mutual labels:  image-retrieval
ImageRetrieval
Content Based Image Retrieval Techniques (e.g. knn, svm using MatLab GUI)
Stars: ✭ 51 (+264.29%)
Mutual labels:  image-retrieval
img classification deep learning
No description or website provided.
Stars: ✭ 19 (+35.71%)
Mutual labels:  image-retrieval
Landmark-Retrival
Easy Landmark Image Recognition with TensorFlow Hub DELF Module
Stars: ✭ 20 (+42.86%)
Mutual labels:  image-retrieval
Image-Retrieval
Image retrieval program made in Tensorflow supporting VGG16, VGG19, InceptionV3 and InceptionV4 pretrained networks and own trained Convolutional autoencoder.
Stars: ✭ 56 (+300%)
Mutual labels:  image-retrieval
FAIRY
Fast and scalable search of whole-slide images via self-supervised deep learning - Nature Biomedical Engineering
Stars: ✭ 43 (+207.14%)
Mutual labels:  image-retrieval
carbon-footprint
Calculate your carbon footprint 🏭👣 from food, transport, purchases, fashion, electricity and digital activities like streaming, NFT or blockchain.
Stars: ✭ 59 (+321.43%)
Mutual labels:  fashion
pytorch-siamese-triplet
One-Shot Learning with Triplet CNNs in Pytorch
Stars: ✭ 74 (+428.57%)
Mutual labels:  triplet-loss

Street2Shop implementation with keras

Task

Clothing image retrieval from consumer(street) image to shop(online) image.

Goal

Compare Models for clothing retrieval. Main: VGG16 + Triplet Loss/Improved Triplet Loss

  1. SingleNetwork for one input
  2. DualNetwork for one input

Dataset

I used DeepFashion Dataset to train and test my model.

  • anchor = street image
  • positive = shop image consistent to anchor
  • negative = shop image not consistent to anchor

data example

Retrieval Sample

T_Shirt

Dress

Coat

Pants

Implementation

These notebooks are implemented by python3.5 or 3.6, keras 2.2.1. Sorry, I use Japanese for comment in my notebook, haha. Please execute this command for building my environment. I trained models for one category respectively.

$ conda env create -n {new environment name} -f env_zkk.yml

Reference

notion

If you encount the error like below, please Quite jupyter process.

Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.

If you close tab without Quit, the jupyter-process would be zombie process. GPU would be occupied by that process and you can't initialize a new process.

reference(https://qiita.com/Ka-k/items/cb942855ab669ff60630)

 TensorFlow-GPUは、デフォルトだとGPUの空いている領域をすべて確保しようとする。すべてです。ALL。  つまり、TensorFlow-GPUを使った機械学習プログラムを複数同時に走らせると1つめは普通に通るけど2つめはGPUを確保できないので初期化に失敗する。これがcuDNN failed to initializeの正体。  そしてJupyterをタブ閉じで終わらせるとGPUを確保したままゾンビ化する。その後Jupyterを立ち上げ直してもGPUはゾンビ化したコードに専有されたままなのでこれも初期化失敗する。これが1回目でも学習失敗する症状の正体。  試しに1つめのコードを走らせたあと、Jupyter上でコードを停止させてから2つめを走らせると、普通に動いた。CUDAもcuDNNもAnacondaも悪くない。悪いのはGPUをひとり占めしようとするTensorFlowとタブを消したらゾンビ化するJupyterやったんや。

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].