This repository is the official implementation of the paper "Representing Prior Knowledge Using Randomly, Weighted Feature Networks for Visual Relationship Detection" . All source codes were implemented in Python 2.7.

The single-hidden-layer Randomly Weighted Feature Network (RWFN) introduced by Hong and Pavlic (2021) was developed as an alternative to neural tensor network approaches for relational learning tasks. Its relatively small footprint combined with the use of two randomized input projections -- an insect-brain-inspired input representation and random Fourier features -- allow it to achieve rich expressiveness for relational learning with relatively low training cost. In particular, when Hong and Pavlic compared RWFN to Logic Tensor Networks (LTNs) for Semantic Image Interpretation (SII) tasks to extract structured semantic descriptions from images, they showed that the RWFN integration of the two hidden, randomized representations better captures relationships among inputs with a faster training process even though it uses far fewer learnable parameters. In this paper, we use RWFNs to perform Visual Relationship Detection (VRD) tasks, which are more challenging SII tasks. A zero-shot learning approach is used with RWFN that can exploit similarities with other seen relationships and background knowledge -- expressed with logical constraints between subjects, relations, and objects -- to achieve the ability to predict triples that do not appear in the training set. The experiments on the Visual Relationship Dataset to compare the performance between RWFNs and LTNs, one of the leading Statistical Relational Learning frameworks, show that RWFNs outperform LTNs for the predicate-detection task while using fewer number of adaptable parameters (1:56 ratio). Furthermore, background knowledge represented by RWFNs can be used to alleviate the incompleteness of training sets even though the space complexity of RWFNs is much smaller than LTNs (1:27 ratio).

To install requirements:

pip install -r requirements.txt

We directly compare the performances between our method and LTNs using Visual Relationship Detection (VRD) dataset. Original source codes of LTNs and dataset are available here. All the details of best hyperparameters for RWFNs are described in the paper.

To run a train use the following command:

python train.py

The trained grounded theories are saved in the models folder in the files; KB_nc_10000.ckpt (no constraints) and KB_wc_10000.ckpt (with constraints) for LTNs and RWFN_KB_nc_10000.ckpt (no constraints) and RWFN_KB_wc_10000.ckpt (with constraints) for RWFNs. The number in the filename (10000) is a parameter in the code to set the number of iterations.

To run the evaluation use the following commands

python predicate_detection.py
python relationship_phrase_detection.py

Then, launch Matlab, move into the Visual-Relationship-Detection-master folder, execute the scripts predicate_detection_RWFN.m and relationship_phrase_detection_RWFN.m and see the results.

All content in this repository is licensed under the MIT license.

You can cite our work:

@inproceedings{hong2022representing,
  title={Representing Prior Knowledge Using Randomly, Weighted Feature Networks for Visual Relationship Detection},
  author={Hong, Jinyung and Pavlic, Ted},
  booktitle={The First International Workshop on Combining Learning and Reasoning: Programming Languages, Formalisms, and Representations~(CLeaR) at the AAAI Conference on Artificial Intelligence},
  year={2022}
}