cnnimageretrieval-pytorch

CNN Image Retrieval in PyTorch: Training and evaluating CNNs for Image Retrieval in PyTorch

This is a Python toolbox that implements the training and testing of the approach described in our papers:

Fine-tuning CNN Image Retrieval with No Human Annotation,
Radenović F., Tolias G., Chum O., TPAMI 2018 [arXiv]

CNN Image Retrieval Learns from BoW: Unsupervised Fine-Tuning with Hard Examples,
Radenović F., Tolias G., Chum O., ECCV 2016 [arXiv]

What is it?

This code implements:

1. Training (fine-tuning) CNN for image retrieval
2. Learning supervised whitening, as post-processing, for global image descriptors
3. Testing CNN image retrieval on Oxford and Paris datasets

Prerequisites

In order to run this toolbox you will need:

1. Python3 (tested with Python 3.7.0 on Debian 8.1)
2. PyTorch deep learning framework (tested with version 1.0.0)
3. All the rest (data + networks) is automatically downloaded with our scripts

Usage

Navigate (cd) to the root of the toolbox [YOUR_CIRTORCH_ROOT]. You can install package with pip3 install . if you need. Make sure to have desired PyTorch and torchvision packages installed.

Example training script is located in YOUR_CIRTORCH_ROOT/cirtorch/examples/train.py

…

For detailed explanation of the options run:

python3 -m cirtorch.examples.train -h

Note: Data and networks used for training and testing are automatically downloaded when using the example script.

Example testing script is located in YOUR_CIRTORCH_ROOT/cirtorch/examples/test.py

python3 -m cirtorch.examples.test [-h] (--network-path NETWORK | --network-offtheshelf NETWORK)
               [--datasets DATASETS] [--image-size N]
               [--multiscale MULTISCALE] [--whitening WHITENING] [--gpu-id N]

For detailed explanation of the options run:

python3 -m cirtorch.examples.test -h

Note: Data used for testing are automatically downloaded when using the example script.

Papers implementation

For example, to train our best network described in the TPAMI 2018 paper run the following command. After each epoch, the fine-tuned network will be tested on the revisited Oxford and Paris benchmarks:

python3 -m cirtorch.examples.train YOUR_EXPORT_DIR --gpu-id '0' --training-dataset 'retrieval-SfM-120k' 
            --test-datasets 'roxford5k,rparis6k' --arch 'resnet101' --pool 'gem' --loss 'contrastive' 
            --loss-margin 0.85 --optimizer 'adam' --lr 5e-7 --neg-num 5 --query-size=2000 
            --pool-size=22000 --batch-size 5 --image-size 362

Networks can be evaluated with learned whitening after each epoch (whitening is estimated at the end of the epoch). To achieve this run the following command. Note that this will significantly slow down the entire training procedure, and you can evaluate networks with learned whitening later on using the example test script.

python3 -m cirtorch.examples.train YOUR_EXPORT_DIR --gpu-id '0' --training-dataset 'retrieval-SfM-120k' 
            --test-datasets 'roxford5k,rparis6k' --test-whiten 'retrieval-SfM-30k' 
            --arch 'resnet101' --pool 'gem' --loss 'contrastive' --loss-margin 0.85 
            --optimizer 'adam' --lr 5e-7 --neg-num 5 --query-size=2000 --pool-size=22000 
            --batch-size 5 --image-size 362

Note: Adjusted (lower) learning rate is set to achieve similar performance as with MatConvNet and PyTorch-0.3.0 implementation of the training.

Pretrained networks trained using the same parameters as in our TPAMI 2018 paper are provided, with precomputed post-processing whitening step. To evaluate them run:

python3 -m cirtorch.examples.test --gpu-id '0' --network-path 'retrievalSfM120k-resnet101-gem' 
                --datasets 'oxford5k,paris6k,roxford5k,rparis6k' 
                --whitening 'retrieval-SfM-120k'
                --multiscale '[1, 1/2**(1/2), 1/2]'

or

python3 -m cirtorch.examples.test --gpu-id '0' --network-path 'retrievalSfM120k-vgg16-gem' 
                --datasets 'oxford5k,paris6k,roxford5k,rparis6k' 
                --whitening 'retrieval-SfM-120k'
                --multiscale '[1, 1/2**(1/2), 1/2]'

The table below shows the performance comparison of networks trained with this framework and the networks used in the paper which were trained with our CNN Image Retrieval in MatConvNet:

Note (June 2022): We updated download files for Oxford 5k and Paris 6k images to use images with blurred faces as suggested by the original dataset owners. Bear in mind, "experiments have shown that one can use the face-blurred version for benchmarking image retrieval with negligible loss of accuracy".

To evaluate your trained network using single scale and without learning whitening:

python3 -m cirtorch.examples.test --gpu-id '0' --network-path YOUR_NETWORK_PATH 
                --datasets 'oxford5k,paris6k,roxford5k,rparis6k'

To evaluate trained network using multi scale evaluation and with learned whitening as post-processing:

python3 -m cirtorch.examples.test --gpu-id '0' --network-path YOUR_NETWORK_PATH 
                --datasets 'oxford5k,paris6k,roxford5k,rparis6k'
                --whitening 'retrieval-SfM-120k' 
                --multiscale '[1, 1/2**(1/2), 1/2]'

Off-the-shelf networks can be evaluated as well, for example:

python3 -m cirtorch.examples.test --gpu-id '0' --network-offtheshelf 'resnet101-gem'
                --datasets 'oxford5k,paris6k,roxford5k,rparis6k'
                --whitening 'retrieval-SfM-120k' 
                --multiscale '[1, 1/2**(1/2), 1/2]'

Networks with projection (FC) layer after global pooling

An alternative architecture includes a learnable FC (projection) layer after the global pooling. It is important to initialize the parameters of this layer with the result of learned whitening. To train such a setup you should run the following commands (the performance will be evaluated every 5 epochs on roxford5k and rparis6k):

python3 -m cirtorch.examples.train YOUR_EXPORT_DIR --gpu-id '0' --training-dataset 'retrieval-SfM-120k' 
            --loss 'triplet' --loss-margin 0.5 --optimizer 'adam' --lr 1e-6 
            --arch 'resnet50' --pool 'gem' --whitening 
            --neg-num 5 --query-size=2000 --pool-size=20000 
            --batch-size 5 --image-size 1024 --epochs 100 
            --test-datasets 'roxford5k,rparis6k' --test-freq 5 

or

python3 -m cirtorch.examples.train YOUR_EXPORT_DIR --gpu-id '0' --training-dataset 'retrieval-SfM-120k' 
            --loss 'triplet' --loss-margin 0.5 --optimizer 'adam' --lr 5e-7 
            --arch 'resnet101' --pool 'gem' --whitening 
            --neg-num 4 --query-size=2000 --pool-size=20000 
            --batch-size 5 --image-size 1024 --epochs 100 
            --test-datasets 'roxford5k,rparis6k' --test-freq 5 

or

python3 -m cirtorch.examples.train YOUR_EXPORT_DIR --gpu-id '0' --training-dataset 'retrieval-SfM-120k' 
            --loss 'triplet' --loss-margin 0.5 --optimizer 'adam' --lr 5e-7 
            --arch 'resnet152' --pool 'gem' --whitening 
            --neg-num 3 --query-size=2000 --pool-size=20000 
            --batch-size 5 --image-size 900 --epochs 100 
            --test-datasets 'roxford5k,rparis6k' --test-freq 5 

for ResNet50, ResNet101, or ResNet152, respectively.

Implementation details:

• The FC layer is initialized with the result of whitening learned in a supervised manner using our training data and off-the-shelf features.
• The whitening for this FC layer is precomputed for popular architectures and pooling methods, see imageretrievalnet.py#L50 for the full list of precomputed FC layers.
• When this FC layer is added in the fine-tuning procedure, the performance is highest if the images are with a similar high-resolution at train and test time.
• When this FC layer is added, the distribution of pairwise distances changes significantly, so roughly twice larger margin should be used for contrastive loss. In this scenario, triplet loss performs slightly better.
• Additional tunning of hyper-parameters can be performed to achieve higher performance or faster training. Note that, in this example, --neg-num and --image-size hyper-parameters are chosen such that the training can be performed on a single GPU with 16 GB of memory.

Pretrained networks with projection layer are provided, trained both on retrieval-SfM-120k (rSfM120k) and google-landmarks-2018 (gl18) train datasets. For this architecture, there is no need to compute whitening as post-processing step (typically the performance boost is insignificant), although one can do that, as well. For example, multi-scale evaluation of ResNet101 with GeM with projection layer trained on google-landmarks-2018 (gl18) dataset using high-resolution images and a triplet loss, is performed with the following script:

python3 -m cirtorch.examples.test_e2e --gpu-id '0' --network 'gl18-tl-resnet101-gem-w' 
            --datasets 'roxford5k,rparis6k' --multiscale '[1, 2**(1/2), 1/2**(1/2)]'

Multi-scale performance of all available pre-trained networks is given in the following table: