This repo implements paper Look Before You Leap: Learning Landmark Features For One-Stage Visual Grounding CVPR 2021.

The core of this paper is Landmark Convolution.

• python 3.7
• pytorch 10.0
• cuda 10.0
• gcc 4.92 or above

1. Then clone the repo and install dependencies.

   git clone https://github.com/svip-lab/LBYLNet.git
   cd LBYLNet
   pip install requirements.txt 

2. You also need to install our landmark feature convolution:

   cd ext
   git clone https://github.com/hbb1/landmarkconv.git
   cd landmarkconv/lib/layers
   python setup.py install --user

3. We follow dataset structure DMS and FAOA. For convience, we have pack them togather, including ReferitGame, RefCOCO, RefCOCO+, RefCOCOg.

   bash data/refer/download_data.sh ./data/refer

4. download the generated index files and place them in ./data/refer. Available at [Gdrive], [One Drive] .
5. download the pretained model of YOLOv3.

   wget -P ext https://pjreddie.com/media/files/yolov3.weights

By default, we use 2 gpus and batchsize 64 with DDP (distributed data-parallel). We have provided several configurations and training log for reproducing our results. If you want to use different hyperparameters or models, you may create configs for yourself. Here are examples:

• For distributed training with gpus :

  CUDA_VISIBLE_DEVICES=0,1 python train.py lbyl_lstm_referit_batch64  --workers 8 --distributed --world_size 1  --dist_url "tcp://127.0.0.1:60006"

• If you use single gpu or won't use distributed training (make sure to adjust the batchsize in the corresponding config file to match your devices):

  CUDA_VISIBLE_DEVICES=0, python train.py lbyl_lstm_referit_batch64  --workers 8

• For evaluation:

  CUDA_VISIBLE_DEVICES=0, python evaluate.py lbyl_lstm_referit_batch64 --testiter 100 --split val

We provide the our retrained models with this re-organized codebase and provide their checkpoints and logs for reproducing the results. To use our trained models, download them from the [Gdrive] and save them into directory cache. Then the file path is expected to be <LBYLNet dir>/cache/nnet/<config>/<dataset>/<config>_100.pkl

Notice: The reproduced performances are occassionally higher or lower (within a reasonable range) than the results reported in the paper.

In this repo, we provide the peformance of our LBYL-Nets below. You can also find the details on <LBYLNet dir>/results and <LBYLNet dir>/logs.

• Performance on ReferitGame (Pr@0.5).

  Dataset	Langauge	Split	Papar	Reproduce
  ReferitGame	LSTM	test	65.48	65.98
  	BERT	test	67.47	68.48

• Performance on RefCOCO (Pr@0.5).

  Dataset	Langauge	Split	Papar	Reproduce
  RefCOCO	LSTM
  		testA	82.18	82.48
  		testB	71.91	71.76
  	BERT
  		testA	82.91	82.82
  		testB	74.15	72.82

• Performance on RefCOCO+ (Pr@0.5).

  Dataset	Langauge	Split	Papar	Reproduce
  RefCOCO+	LSTM	val	66.64	66.71
  		testA	73.21	72.63
  		testB	56.23	55.88
  	BERT	val	68.64	68.76
  		testA	73.38	73.73
  		testB	59.49	59.62

• Performance on RefCOCOg (Pr@0.5).

  Dataset	Langauge	Split	Papar	Reproduce
  RefCOCOg	LSTM	val	58.72	60.03
  	BERT	val	62.70	63.20

We also provide demo scripts to test if the repo is corretly installed. After installing the repo and download the pretained weights, you should be able to use the LBYL-Net to ground your own images.

python demo.py

you can change the model, image or phrase in the demo.py. You will see the output image in imgs/demo_out.jpg.

#!/usr/bin/env python
import cv2
import torch
from core.test.test import _visualize
from core.groundors import Net 
# pick one model
cfg_file = "lbyl_bert_unc+_batch64"
detector = Net(cfg_file, iter=100)
# inference
image = cv2.imread('imgs/demo.jpeg')
phrase = 'the green gaint'
bbox = detector(image, phrase)
_visualize(image, pred_bbox=bbox, phrase=phrase, save_path='imgs/demo_out.jpg', color=(1, 174, 245), draw_phrase=True)

Input:

Output:

This repo is organized as CornerNet-Lite and the code is partially from FAOA (e.g. data preparation) and MAttNet (e.g. LSTM). We thank for their great works.

If you use any part of this repo in your research, please cite our paper:

@InProceedings{huang2021look,
      title={Look Before You Leap: Learning Landmark Features for One-Stage Visual Grounding}, 
      author={Huang, Binbin and Lian, Dongze and Luo, Weixin and Gao, Shenghua},
      booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
      month = {June},
      year={2021},
}