VD-BERT: A Unified Vision and Dialog Transformer with BERT

Related tags

Deep LearningVD-BERT
Overview

VD-BERT: A Unified Vision and Dialog Transformer with BERT

PyTorch Code for the following paper at EMNLP2020:
Title: VD-BERT: A Unified Vision and Dialog Transformer with BERT [pdf]
Authors: Yue Wang, Shafiq Joty, Michael R. Lyu, Irwin King, Caiming Xiong, Steven C.H. Hoi
Institute: Salesforce Research and CUHK
Abstract
Visual dialog is a challenging vision-language task, where a dialog agent needs to answer a series of questions through reasoning on the image content and dialog history. Prior work has mostly focused on various attention mechanisms to model such intricate interactions. By contrast, in this work, we propose VD-BERT, a simple yet effective framework of unified vision-dialog Transformer that leverages the pretrained BERT language models for Visual Dialog tasks. The model is unified in that (1) it captures all the interactions between the image and the multi-turn dialog using a single-stream Transformer encoder, and (2) it supports both answer ranking and answer generation seamlessly through the same architecture. More crucially, we adapt BERT for the effective fusion of vision and dialog contents via visually grounded training. Without the need of pretraining on external vision-language data, our model yields new state of the art, achieving the top position in both single-model and ensemble settings (74.54 and 75.35 NDCG scores) on the visual dialog leaderboard.

Framework illustration
VD-BERT framework

Installation

Package: Pytorch 1.1; We alo provide our Dockerfile and YAML file for setting up experiments in Google Cloud Platform (GCP).
Data: you can obtain the VisDial data from here
Visual features: we provide bottom-up attention visual features of VisDial v1.0 on data/img_feats1.0/. If you would like to extract visual features for other images, please refer to this docker image. We provide the running script on data/visual_extract_code.py, which should be used inside the provided bottom-up-attention image.

Code explanation

vdbert: store the main training and testing python files, data loader code, metrics and the ensemble code;

pytorch_pretrained_bert: mainly borrow from the Huggingface's pytorch-transformers v0.4.0;

  • modeling.py: we modify or add two classes: BertForPreTrainingLossMask and BertForVisDialGen;
  • rank_loss.py: three ranking methods: ListNet, ListMLE, approxNDCG;

sh: shell scripts to run the experiments

pred: store two json files for best single-model (74.54 NDCG) and ensemble model (75.35 NDCG)

model: You can download a pretrained model from https://storage.cloud.google.com/sfr-vd-bert-research/v1.0_from_BERT_e30.bin

Running experiments

Below the running example scripts for pretraining, finetuning (including dense annotation), and testing.

  • Pretraining bash sh/pretrain_v1.0_mlm_nsp_g4.sh
  • Finetuning for discriminative bash sh/finetune_v1.0_disc_g4.sh
  • Finetuning for discriminative specifically on dense annotation bash sh/finetune_v1.0_disc_dense_g4.sh
  • Finetuning for generative bash sh/finetune_v1.0_gen_g4.sh
  • Testing for discriminative on validation bash sh/test_v1.0_disc_val.sh
  • Testing for generative on validation bash sh/test_v1.0_gen_val.sh
  • Testing for discriminative on test bash sh/test_v1.0_disc_test.sh

Notation: mlm: masked language modeling, nsp: next sentence prediction, disc: discriminative, gen: generative, g4: 4 gpus, dense: dense annotation

Citation

If you find the code useful in your research, please consider citing our paper:

@inproceedings{
    wang2020vdbert,
    title={VD-BERT: A Unified Vision and Dialog Transformer with BERT},
    author={Yue Wang, Shafiq Joty, Michael R. Lyu, Irwin King, Caiming Xiong, Steven C.H. Hoi},
    booktitle={Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, EMNLP 2020},
    year={2020},
}

License

This project is licensed under the terms of the MIT license.

Owner
Salesforce
A variety of vendor agnostic projects which power Salesforce
Salesforce
Benchmark tools for Compressive LiDAR-to-map registration

Benchmark tools for Compressive LiDAR-to-map registration This repo contains the released version of code and datasets used for our IROS 2021 paper: "

Allie 9 Nov 24, 2022
SMIS - Semantically Multi-modal Image Synthesis(CVPR 2020)

Semantically Multi-modal Image Synthesis Project page / Paper / Demo Semantically Multi-modal Image Synthesis(CVPR2020). Zhen Zhu, Zhiliang Xu, Anshen

316 Dec 01, 2022
SCU OlympicsRunning Baseline

Competition 1v1 running Environment check details in Jidi Competition RLChina2021智能体竞赛 做出的修改: 奖励重塑:修改了环境,重新设置了奖励的分配,使得奖励组成不只有零和博弈,还有探索环境的奖励。 算法微调:修改了官

ZiSeoi Wong 2 Nov 23, 2021
Visual Question Answering in Pytorch

Visual Question Answering in pytorch /!\ New version of pytorch for VQA available here: https://github.com/Cadene/block.bootstrap.pytorch This repo wa

Remi 672 Jan 01, 2023
This framework implements the data poisoning method found in the paper Adversarial Examples Make Strong Poisons

Adversarial poison generation and evaluation. This framework implements the data poisoning method found in the paper Adversarial Examples Make Strong

31 Nov 01, 2022
Point-NeRF: Point-based Neural Radiance Fields

Point-NeRF: Point-based Neural Radiance Fields Project Sites | Paper | Primary c

Qiangeng Xu 662 Jan 01, 2023
Official Pytorch Implementation of Adversarial Instance Augmentation for Building Change Detection in Remote Sensing Images.

IAug_CDNet Official Implementation of Adversarial Instance Augmentation for Building Change Detection in Remote Sensing Images. Overview We propose a

53 Dec 02, 2022
Sentinel-1 vessel detection model used in the xView3 challenge

sar_vessel_detect Code for the AI2 Skylight team's submission in the xView3 competition (https://iuu.xview.us) for vessel detection in Sentinel-1 SAR

AI2 6 Sep 10, 2022
Simple reference implementation of GraphSAGE.

Reference PyTorch GraphSAGE Implementation Author: William L. Hamilton Basic reference PyTorch implementation of GraphSAGE. This reference implementat

William L Hamilton 861 Jan 06, 2023
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
This is an open source library implementing hyperbox-based machine learning algorithms

hyperbox-brain is a Python open source toolbox implementing hyperbox-based machine learning algorithms built on top of scikit-learn and is distributed

Complex Adaptive Systems (CAS) Lab - University of Technology Sydney 21 Dec 14, 2022
3D ResNet Video Classification accelerated by TensorRT

Activity Recognition TensorRT Perform video classification using 3D ResNets trained on Kinetics-400 dataset and accelerated with TensorRT P.S Click on

Akash James 39 Nov 21, 2022
An implementation of Equivariant e2 convolutional kernals into a convolutional self attention network, applied to radio astronomy data.

EquivariantSelfAttention An implementation of Equivariant e2 convolutional kernals into a convolutional self attention network, applied to radio astro

2 Nov 09, 2021
Implementation of Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning

advantage-weighted-regression Implementation of Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning, by Peng et al. (

Omar D. Domingues 1 Dec 02, 2021
Course about deep learning for computer vision and graphics co-developed by YSDA and Skoltech.

Deep Vision and Graphics This repo supplements course "Deep Vision and Graphics" taught at YSDA @fall'21. The course is the successor of "Deep Learnin

Yandex School of Data Analysis 160 Jan 02, 2023
A Python implementation of the Locality Preserving Matching (LPM) method for pruning outliers in image matching.

LPM_Python A Python implementation of the Locality Preserving Matching (LPM) method for pruning outliers in image matching. The code is established ac

AoxiangFan 11 Nov 07, 2022
A knowledge base construction engine for richly formatted data

Fonduer is a Python package and framework for building knowledge base construction (KBC) applications from richly formatted data. Note that Fonduer is

HazyResearch 386 Dec 05, 2022
Official implementation of "Can You Spot the Chameleon? Adversarially Camouflaging Images from Co-Salient Object Detection" in CVPR 2022.

Jadena Official implementation of "Can You Spot the Chameleon? Adversarially Camouflaging Images from Co-Salient Object Detection" in CVPR 2022. arXiv

Qing Guo 13 Nov 29, 2022
VOS: Learning What You Don’t Know by Virtual Outlier Synthesis

VOS This is the source code accompanying the paper VOS: Learning What You Don’t

248 Dec 25, 2022
RM Operation can equivalently convert ResNet to VGG, which is better for pruning; and can help RepVGG perform better when the depth is large.

RMNet: Equivalently Removing Residual Connection from Networks This repository is the official implementation of "RMNet: Equivalently Removing Residua

184 Jan 04, 2023