Distance-Ratio-Based Formulation for Metric Learning

Overview

Distance-Ratio-Based Formulation for Metric Learning

Environment

Preparing datasets

CUB

  • Change directory to /filelists/CUB
  • run source ./download_CUB.sh

One might need to manually download CUB data from http://www.vision.caltech.edu/visipedia-data/CUB-200-2011/CUB_200_2011.tgz.

mini-ImageNet

  • Change directory to /filelists/miniImagenet
  • run source ./download_miniImagenet.sh (WARNING: This would download the 155G ImageNet dataset.)

To only download 'miniImageNet dataset' and not the whole 155G ImageNet dataset:

(Download 'csv' files from the codes in /filelists/miniImagenet/download_miniImagenet.sh. Then, do the following.)

First, download zip file from https://drive.google.com/file/d/0B3Irx3uQNoBMQ1FlNXJsZUdYWEE/view (It is from https://github.com/oscarknagg/few-shot). After unzipping the zip file at /filelists/miniImagenet, run a script /filelists/miniImagenet/prepare_mini_imagenet.py which is modified from https://github.com/oscarknagg/few-shot/blob/master/scripts/prepare_mini_imagenet.py. Then, run /filelists/miniImagenet/write_miniImagenet_filelist2.py.

Train

Run python ./train.py --dataset [DATASETNAME] --model [BACKBONENAME] --method [METHODNAME] --train_aug [--OPTIONARG]

To also save training analyses results, for example, run python ./train.py --dataset miniImagenet --model Conv4 --method protonet_S --train_aug --n_shot 5 --train_n_way 5 --test_n_way 5 > record/miniImagenet_Conv4_proto_S_5s5w.txt

train_models.ipynb contains codes for our experiments.

Save features

Save the extracted feature before the classifaction layer to increase test speed.

For instance, run python ./save_features.py --dataset miniImagenet --model Conv4 --method protonet_S --train_aug --n_shot 5 --train_n_way 5

Test

For example, run python ./test.py --dataset miniImagenet --model Conv4 --method protonet_S --train_aug --n_shot 5 --train_n_way 5 --test_n_way 5

Analyze training

Run /record/analyze_training_1shot.ipynb and /record/analyze_training_5shot.ipynb to analyze training results (norm ratio, con-alpha ratio, div-alpha ratio, and con-div ratio)

Results

The test results will be recorded in ./record/results.txt

Visual comparison of softmax-based and distance-ratio-based (DR) formulation

The following images visualize confidence scores of red class when the three points are the representing points of red, green, and blue classes.

Softmax-based formulation DR formulation

References and licence

Our repository (a set of codes) is forked from an original repository (https://github.com/wyharveychen/CloserLookFewShot) and codes are under the same licence (LICENSE.txt) as the original repository except for the following.

/filelists/miniImagenet/prepare_mini_imagenet.py file is modifed from https://github.com/oscarknagg/few-shot. It is under a different licence in /filelists/miniImagenet/prepare_mini_imagenet.LICENSE

Copyright and licence notes (including the copyright note in /data/additional_transforms.py) are from the original repositories (https://github.com/wyharveychen/CloserLookFewShot and https://github.com/oscarknagg/few-shot).

Modifications

List of modified or added files (or folders) compared to the original repository (https://github.com/wyharveychen/CloserLookFewShot):

io_utils.py backbone.py configs.py train.py save_features.py test.py utils.py README.md train_models.ipynb /methods/__init__.py /methods/protonet_S.py /methods/meta_template.py /methods/protonet_DR.py /methods/softmax_1nn.py /methods/DR_1nn.py /models/ /filelists/miniImagenet/prepare_mini_imagenet.py /filelists/miniImagenet/prepare_mini_imagenet.LICENSE /filelists/miniImagenet/write_miniImagenet_filelist2.py /record/ /record/preprocessed/ /record/analyze_training_1shot.ipynb /record/analyze_training_5shot.ipynb

My (Hyeongji Kim) main contributions (modifications) are in /methods/meta_template.py, /methods/protonet_DR.py, /methods/softmax_1nn.py, /methods/DR_1nn.py, /record/analyze_training_1shot.ipynb, and /record/analyze_training_5shot.ipynb.

Owner
Hyeongji Kim
Hyeongji Kim
Localization Distillation for Object Detection

Localization Distillation for Object Detection This repo is based on mmDetection. This is the code for our paper: Localization Distillation

274 Dec 26, 2022
Augmented Traffic Control: A tool to simulate network conditions

Augmented Traffic Control Full documentation for the project is available at http://facebook.github.io/augmented-traffic-control/. Overview Augmented

Meta Archive 4.3k Jan 08, 2023
SeisComP/SeisBench interface to enable deep-learning (re)picking in SeisComP

scdlpicker SeisComP/SeisBench interface to enable deep-learning (re)picking in SeisComP Objective This is a simple deep learning (DL) repicker module

Joachim Saul 6 May 13, 2022
Semantic Segmentation for Aerial Imagery using Convolutional Neural Network

This repo has been deprecated because whole things are re-implemented by using Chainer and I did refactoring for many codes. So please check this newe

Shunta Saito 27 Sep 23, 2022
Pytorch implementation of the unsupervised object discovery method LOST.

LOST Pytorch implementation of the unsupervised object discovery method LOST. More details can be found in the paper: Localizing Objects with Self-Sup

Valeo.ai 189 Dec 25, 2022
This repository provides a PyTorch implementation and model weights for HCSC (Hierarchical Contrastive Selective Coding)

HCSC: Hierarchical Contrastive Selective Coding This repository provides a PyTorch implementation and model weights for HCSC (Hierarchical Contrastive

YUANFAN GUO 111 Dec 20, 2022
Learning Multiresolution Matrix Factorization and its Wavelet Networks on Graphs

Project Learning Multiresolution Matrix Factorization and its Wavelet Networks on Graphs, https://arxiv.org/pdf/2111.01940.pdf. Authors Truong Son Hy

5 Jun 28, 2022
Train a state-of-the-art yolov3 object detector from scratch!

TrainYourOwnYOLO: Building a Custom Object Detector from Scratch This repo let's you train a custom image detector using the state-of-the-art YOLOv3 c

AntonMu 616 Jan 08, 2023
PyTorch implementation for View-Guided Point Cloud Completion

PyTorch implementation for View-Guided Point Cloud Completion

22 Jan 04, 2023
PyTorch implementation of paper: HPNet: Deep Primitive Segmentation Using Hybrid Representations.

HPNet This repository contains the PyTorch implementation of paper: HPNet: Deep Primitive Segmentation Using Hybrid Representations. Installation The

Siming Yan 42 Dec 07, 2022
v objective diffusion inference code for JAX.

v-diffusion-jax v objective diffusion inference code for JAX, by Katherine Crowson (@RiversHaveWings) and Chainbreakers AI (@jd_pressman). The models

Katherine Crowson 186 Dec 21, 2022
Official Pytorch implementation of "CLIPstyler:Image Style Transfer with a Single Text Condition"

CLIPstyler Official Pytorch implementation of "CLIPstyler:Image Style Transfer with a Single Text Condition" Environment Pytorch 1.7.1, Python 3.6 $ c

201 Dec 29, 2022
Emblaze - Interactive Embedding Comparison

Emblaze - Interactive Embedding Comparison Emblaze is a Jupyter notebook widget for visually comparing embeddings using animated scatter plots. It bun

CMU Data Interaction Group 77 Nov 24, 2022
The 3rd place solution for competition

The 3rd place solution for competition "Lyft Motion Prediction for Autonomous Vehicles" at Kaggle Team behind this solution: Artsiom Sanakoyeu [Homepa

Artsiom 104 Nov 22, 2022
Implementation of E(n)-Transformer, which extends the ideas of Welling's E(n)-Equivariant Graph Neural Network to attention

E(n)-Equivariant Transformer (wip) Implementation of E(n)-Equivariant Transformer, which extends the ideas from Welling's E(n)-Equivariant G

Phil Wang 132 Jan 02, 2023
Pytorch implementation of face attention network

Face Attention Network Pytorch implementation of face attention network as described in Face Attention Network: An Effective Face Detector for the Occ

Hooks 312 Dec 09, 2022
Generating Radiology Reports via Memory-driven Transformer

R2Gen This is the implementation of Generating Radiology Reports via Memory-driven Transformer at EMNLP-2020. Citations If you use or extend our work,

CUHK-SZ NLP Group 101 Dec 13, 2022
Benchmark for evaluating open-ended generation

OpenMEVA Contributed by Jian Guan, Zhexin Zhang. Thank Jiaxin Wen for DeBugging. OpenMEVA is a benchmark for evaluating open-ended story generation me

25 Nov 15, 2022
Selective Wavelet Attention Learning for Single Image Deraining

SWAL Code for Paper "Selective Wavelet Attention Learning for Single Image Deraining" Prerequisites Python 3 PyTorch Models We provide the models trai

Bobo 9 Jun 17, 2022