Pith. sign in

REVIEW 15 cited by

Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1610.02391 v4 pith:MFU6MCSS submitted 2016-10-07 cs.CV cs.AIcs.LG

Ramprasaath R. Selvaraju , Michael Cogswell , Abhishek Das , Ramakrishna Vedantam , Devi Parikh , Dhruv Batra This is my paper
classification cs.CVcs.AIcs.LG
keywords grad-cammodelscnnsexplanationsimagelocalizationvisualcaptioning
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We propose a technique for producing "visual explanations" for decisions from a large class of CNN-based models, making them more transparent. Our approach - Gradient-weighted Class Activation Mapping (Grad-CAM), uses the gradients of any target concept, flowing into the final convolutional layer to produce a coarse localization map highlighting important regions in the image for predicting the concept. Grad-CAM is applicable to a wide variety of CNN model-families: (1) CNNs with fully-connected layers, (2) CNNs used for structured outputs, (3) CNNs used in tasks with multimodal inputs or reinforcement learning, without any architectural changes or re-training. We combine Grad-CAM with fine-grained visualizations to create a high-resolution class-discriminative visualization and apply it to off-the-shelf image classification, captioning, and visual question answering (VQA) models, including ResNet-based architectures. In the context of image classification models, our visualizations (a) lend insights into their failure modes, (b) are robust to adversarial images, (c) outperform previous methods on localization, (d) are more faithful to the underlying model and (e) help achieve generalization by identifying dataset bias. For captioning and VQA, we show that even non-attention based models can localize inputs. We devise a way to identify important neurons through Grad-CAM and combine it with neuron names to provide textual explanations for model decisions. Finally, we design and conduct human studies to measure if Grad-CAM helps users establish appropriate trust in predictions from models and show that Grad-CAM helps untrained users successfully discern a 'stronger' nodel from a 'weaker' one even when both make identical predictions. Our code is available at https://github.com/ramprs/grad-cam/, along with a demo at http://gradcam.cloudcv.org, and a video at youtu.be/COjUB9Izk6E.

Discussion (0). Sign in to comment.

Forward citations

Cited by 15 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Towards Verified and Targeted Explanations through Formal Methods

    cs.LG 2026-04 accept novelty 7.0 of 10

    ViTaX certifies targeted semifactual robustness: a minimal feature subset can be perturbed by ε without flipping a neural network from class y to a user-specified high-risk class t.

  2. Simulation-based inference using splitting schemes for partially observed diffusions in chemical reaction networks

    stat.ME 2025-08 unverdicted novelty 6.0 of 10

    Chemical Langevin equations are rewritten as perturbed CIR-type SDEs, enabling a structure-preserving splitting scheme and an ABC-SMC algorithm for inference on partially observed reaction networks.

  3. On Spectral Properties of Gradient-based Explanation Methods

    cs.LG 2025-08 conditional novelty 6.0 of 10

    Gradient-based explanations behave like frequency-band selectors: the gradient acts as a high-pass filter, perturbation as a low-pass filter, and their combination creates explanations that shift with the perturbation scale.

  4. Synthetic Generation and Latent Projection Denoising of Rim Lesions in Multiple Sclerosis

    eess.IV 2025-05 conditional novelty 6.0 of 10

    Synthetic GAN images plus latent projection denoising of ambiguous rim lesions improve automated rim lesion detection on QSM.

  5. KIGNet: Physics-Motivated Multi-Graph Representation Learning for Explainable Jet Tagging

    hep-ph 2025-12 conditional novelty 5.0 of 10

    E-PCN reaches 94.67% macro-accuracy on 10-class jet tagging by weighting graphs with angular separation, transverse momentum, momentum fraction, and invariant mass, with Grad-CAM showing the first two account for 76% ...

  6. Systematic Evaluation of Attribution Methods: Eliminating Threshold Bias and Revealing Method-Dependent Performance Patterns

    cs.LG 2025-09 conditional novelty 5.0 of 10

    Averaging IoU across many thresholds ranks XRAI above LIME and Integrated Gradients variants on HAM10000, and shows single-threshold attribution rankings are unstable.

  7. An Explainable Diagnostic Framework for Neurodegenerative Dementias via Reinforcement-Optimized LLM Reasoning

    cs.LG 2025-05 conditional novelty 5.0 of 10

    A GRPO-tuned 8B LLM diagnoses Alzheimer's and frontotemporal dementia subtypes from synthetic MRI-based radiology reports with accuracy comparable to a 3D vision transformer.

  8. PulmoSight-XAI: An Explainable Multi-View Attention Ensemble with Gradient Boosting Meta-Learning for Multi-Label Chest X-Ray Classification

    cs.CV 2026-07 conditional novelty 4.0 of 10

    View-specific multi-scale CBAM CNN ensembles plus hybrid ASL/focal loss and two-level gradient-boosting stacking reach ~0.93/0.92 macro AUROC on a CheXpert-style multi-label CXR dataset.

  9. Structured Spectral Graph Representation Learning for Multi-label Abnormality Analysis from 3D CT Scans

    cs.CV 2025-10 conditional novelty 4.0 of 10

    A graph-of-slice-triplets encoder with spectral convolution outperforms 3D CNN/Transformer baselines on multi-label chest CT abnormality classification and transfers to report generation and abdominal CT.

  10. Towards Optimal Convolutional Transfer Learning Architectures for Breast Lesion Classification and ACL Tear Detection

    cs.CV 2025-08 reject novelty 4.0 of 10

    On two public medical imaging tasks, 1D convolutional classifiers with skip connections, ResNet50 backbones, and partial unfreezing gave the best validation AUC, while RadImageNet pretraining did not beat ImageNet.

  11. On the Complexity-Faithfulness Trade-off of Gradient-Based Explanations

    cs.LG 2025-08 reject novelty 4.0 of 10

    The paper introduces EF and ΔEF as spectral metrics, but ΔEF is derived from EF, making the complexity-faithfulness trade-off partly tautological.

  12. Faithful, Interpretable Chest X-ray Diagnosis with Anti-Aliased B-cos Networks

    cs.CV 2025-07 conditional novelty 4.0 of 10

    Combining B-cos networks with anti-aliasing pooling (FLC or BlurPool) reduces grid artifacts in chest X-ray explanation maps while keeping diagnostic accuracy close to baseline networks.

  13. CASE: Contrastive Activation for Saliency Estimation

    cs.CV 2025-06 conditional novelty 4.0 of 10

    CASE removes gradient components shared with confused classes to produce more class-distinct saliency maps, validated on a top-k overlap diagnostic where many existing methods show class-insensitive behavior.

  14. XAI-Guided Analysis of Residual Networks for Interpretable Pneumonia Detection in Paediatric Chest X-rays

    eess.IV 2025-07 conditional novelty 3.0 of 10

    A fine-tuned ResNet-50 with Grad-CAM and Monte Carlo dropout reports 95.94% accuracy and 98.91% AUC for pediatric pneumonia on the Kermany chest X-ray dataset.

  15. A Survey on Training-free Open-Vocabulary Semantic Segmentation

    cs.CV 2025-05 conditional novelty 2.0 of 10

    A structured review of over 30 training-free open-vocabulary semantic segmentation methods, organized by whether they rely on CLIP alone, auxiliary visual foundation models, or generative models.

Pith tools