🤖 Visualize Vision Transformer attribution maps and more!

This repository contains the source code for the new Absolute Magnitude Layer-Wise Relevance Propagation attribution method and the Global Evaluation Metric described in the paper https://dl.acm.org/doi/10.1145/3649458 .

🔎 Absolute Magnitude Layer-Wise Relevance Propagation

A novel Layer-Wise Propagation rule, referred to as Absolute Magnitude Layer-Wise Relevance Propagation (absLRP). This rule effectively addresses the issue of incorrect relative attribution between neurons within the same layer that exhibit varying absolute magnitude activations. We apply this rule to three different architectures, including the very recent Vision Transformer.

Figure 1. absLRP visualizations for Vision Transformer architecture - PascalVOC

Figure 2. absLRP visualizations for VGG architecture - ImageNet

🔬 Usage

Load a model from timm and apply the absLRP rule: ```python

load the model

device = "cuda" model = timm.createmodel("vitbasepatch16224", pretrained=True)

model = timm.create_model("vgg16", pretrained=True)

model = timm.create_model("resnet50", pretrained=True)

model.eval() model.to(device)

apply the absLRP rule to the model

VisionTransformerAbsLRPRule().apply(model)

VGGAbsLRPRule().apply(model)

ResNetAbsLRPRule().apply(model)

Load inference images and preprocess: python isvit = isinstance(model, VisionTransformer) x = torch.stack( [ preprocessimage(Image.open("images/dogcat.jpeg"), isvit), preprocessimage(Image.open("images/hedgehog.jpg"), isvit), ] ) Calculate contrastive relevance using absLRP and visualize: python relevancymethod = AbsLRPRelevancyMethod(model, device) relevance = relevancymethod.relevancy(x) visualizebatch(x, relevance, isvit) ```

📊 Global Evaluation Metric

A new evaluation method, Global Attribution Evaluation (GAE), which offers a novel perspective on evaluating faithfulness and robustness of an attribution method by utilizing gradient-based masking, while combining those results with a localization method to achieve a comprehensive evaluation of explanation quality in a single score.

Figure 3. Top and bottom 5 scoring images on GAE metric out of a randomly sampled 1024 images - absLRP VGG ImageNet

🔬 Usage

Import the required libraries python from abslrp_gae.gae.gae import GlobalEvaluationMetric Define a dictionary of relevancy methods: python relevancy_methods = { "abslrp": relevancy_method, } Run the metric python metric = GlobalEvaluationMetric() metric.run( relevancy_methods=relevancy_methods, model=base_model, # original model dataset=dataset, batch_size=16, ) Plot the results python metric.plot()

📞 Contact

🔗 Citation

Please use the following BibText entry to cite our work: bibtex @article{10.1145/3649458, author = {Vukadin, Davor and Afri\'{c}, Petar and \v{S}ili\'{c}, Marin and Dela\v{c}, Goran}, title = {Advancing Attribution-Based Neural Network Explainability through Relative Absolute Magnitude Layer-Wise Relevance Propagation and Multi-Component Evaluation}, year = {2024}, issue_date = {June 2024}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, volume = {15}, number = {3}, issn = {2157-6904}, url = {https://doi.org/10.1145/3649458}, doi = {10.1145/3649458}, journal = {ACM Trans. Intell. Syst. Technol.}, month = {apr}, articleno = {47}, numpages = {30}, keywords = {Explainable artificial intelligence, Vision Transformer, layer-wise relevance propagation, attribution-based evaluation} }