Recurrent VLN-BERT

Code of the CVPR 2021 Oral paper:
A Recurrent Vision-and-Language BERT for Navigation
Yicong Hong, Qi Wu, Yuankai Qi, Cristian Rodriguez-Opazo, Stephen Gould

[Paper & Appendices] [GitHub]

"Neo : Are you saying I have to choose whether Trinity lives or dies? The Oracle : No, you've already made the choice. Now you have to understand it." --- The Matrix Reloaded (2003).

🎁 Features

Enable DDP during training and inference.
Support the latest version of Matterport3DSimulator.

Prerequisites

Installation

Install the Matterport3D Simulator.

Please find the versions of packages in our environment here.

Install the Pytorch-Transformers. In particular, we use this version (same as OSCAR) in our experiments.

Data Preparation

Please follow the instructions below to prepare the data in directories:

MP3D navigability graphs: connectivity
- Download the connectivity maps [23.8MB].
R2R data: data
- Download the R2R data [5.8MB].
Augmented data: data/prevalent
- Download the collected triplets in PREVALENT [1.5GB] (pre-processed for easy use).
MP3D image features: img_features
- Download the Scene features [4.2GB] (ResNet-152-Places365).

Initial OSCAR and PREVALENT weights

Please refer to vlnbert_init.py to set up the directories.

Pre-trained OSCAR weights
- Download the base-no-labels following this guide.
Pre-trained PREVALENT weights
- Download the pytorch_model.bin from here.

Trained Network Weights

Recurrent-VLN-BERT: snap
- Download the trained network weights [2.5GB] for our OSCAR-based and PREVALENT-based models.

R2R Navigation

Please read Peter Anderson's VLN paper for the R2R Navigation task.

Reproduce Testing Results

To replicate the performance reported in our paper, load the trained network weights and run validation:

bash run/test_agent.bash

The results will be saved to logs/VLNBERT-test-Prevalent/snap/submit_test_unseen.json.

Alternatively, you can run validate model performance locally on the validation sets:

bash run/val_agent.bash

The results will be saved to logs/VLNBERT-val-Prevalent/eval.txt.

You can simply switch between the OSCAR-based and the PREVALENT-based VLN models by changing the arguments vlnbert (oscar or prevalent) and load (trained model paths).

Training

Navigator

To train the network from scratch, simply run:

bash run/train_agent.bash

The trained Navigator will be saved under snap/.

Citation

If you use or discuss our Recurrent VLN-BERT, please cite our paper:

@InProceedings{Hong_2021_CVPR,
    author    = {Hong, Yicong and Wu, Qi and Qi, Yuankai and Rodriguez-Opazo, Cristian and Gould, Stephen},
    title     = {A Recurrent Vision-and-Language BERT for Navigation},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2021},
    pages     = {1643-1653}
}

GengzeZhou/Recurrent-VLN