ZS-F-VQA

Zero-shot Visual Question Answering using Knowledge Graph [ ISWC 2021 ]

In this work, we propose a Zero-shot VQA algorithm using knowledge graphs and a mask-based learning mechanism for better incorporating external knowledge, and present new answer-based Zero-shot VQA splits for the F-VQA dataset.

🔔 News

• 2024-02 We preprint our Survey Knowledge Graphs Meet Multi-Modal Learning: A Comprehensive Survey [Repo].

Model Architecture

Usage

Requirements

• python >= 3.5
• PyTorch >= 1.6.0

For more detail of requirements:

 pip install -r requirements.txt

Data

Location of 5 F-VQA train / test data split:

• data/KG_VQA/fvqa/exp_data/train_data
• data/KG_VQA/fvqa/exp_data/test_data

Location of 5 ZS-F-VQA train / test data split:

• data/KG_VQA/fvqa/exp_data/train_seen_data
• data/KG_VQA/fvqa/exp_data/test_unseen_data

Answers are available at data/KG_VQA/data/FVQA/new_dataset_release/.

Image:

• Image folder (put all your .JPEG/.jpg file here): data/KG_VQA/fvqa/exp_data/images/images
• Image feature:
  • fvqa-resnet-14x14.h5 pretrained: GoogleDrive or BaiduCloud (password:16vd)
  • fvqa36_imgid2idx.pkl and fvqa_36.hdf5 pretrained: GoogleDrive or BaiduCloud (password:zsqa)
• Origin images are available at FVQA with download_link.
• Other vqa dataset: you could generate a pretrained image feature via this way (Guidance / code)
• The generated .h file should be placed in : data/KG_VQA/fvqa/exp_data/common_data/.

Answer / Qusetion vocab:

• The generated file answer.vocab.fvqa.json & question.vocab.fvqa.json now are available at : data/KG_VQA/fvqa/exp_data/common_data/.
• Other vqa dataset: code for process answer vocab and process questions vocab

---

Pretrained Model (url)

Download it and overwrite data/KG_VQA/fvqa/model_save

Parameter

[--KGE {TransE,ComplEx,TransR,DistMult}] [--KGE_init KGE_INIT] [--GAE_init GAE_INIT] [--ZSL ZSL] [--entity_num {all,4302}] [--data_choice {0,1,2,3,4}]
               [--name NAME] [--no-tensorboard] --exp_name EXP_NAME [--dump_path DUMP_PATH] [--exp_id EXP_ID] [--random_seed RANDOM_SEED] [--freeze_w2v {0,1}]
               [--ans_net_lay {0,1,2}] [--fact_map {0,1}] [--relation_map {0,1}] [--now_test {0,1}] [--save_model {0,1}] [--joint_test_way {0,1}] [--top_rel TOP_REL]
               [--top_fact TOP_FACT] [--soft_score SOFT_SCORE] [--mrr MRR]

Available model for training: Up-Down, BAN, SAN, MLP

You can try your own model via adding it (.py) to : main/code/model/.

For more details: code/config.py

---

Running

cd code

For data check:

• python deal_data.py --exp_name data_check

General VQA:

• train: bash run_FVQA_train.sh
• test: bash run_FVQA.sh

ZSL/GZSL VQA:

• train: bash run_ZSL_train.sh
• test: bash run_ZSL.sh

Note:

• you can open the .sh file for parameter modification.

Result:

• Log file will be saved to: code/dump
• model will be saved to: data/KG_VQA/fvqa/model_save

Explainable

Acknowledgements

Thanks for the following released works:

SciencePlots, ramen, GAE, vqa-winner-cvprw-2017, faster-rcnn, VQA, BAN, commonsense-kg-completion, bottom-up-attention-vqa, FVQA, answer_embedding, torchlight

Cite:

Please condiser citing this paper if you use the code

@inproceedings{chen2021zero,
  title={Zero-Shot Visual Question Answering Using Knowledge Graph},
  author={Chen, Zhuo and Chen, Jiaoyan and Geng, Yuxia and Pan, Jeff Z and Yuan, Zonggang and Chen, Huajun},
  booktitle={International Semantic Web Conference},
  pages={146--162},
  year={2021},
  organization={Springer}
}

For more details, please submit a issue or contact Zhuo Chen.