Rethinking Global Average Pooling: Your Classifier Is Secretly a Multi-Instance Learner

• URL: http://arxiv.org/abs/2606.14555v1
• Date: Fri, 12 Jun 2026 15:35:05 GMT
• Title: Rethinking Global Average Pooling: Your Classifier Is Secretly a Multi-Instance Learner
• Abstract summary: Modern image classifiers widely adopt global average pooling (GAP) followed by a linear classification head.<n>This linearity ensures that the image-level logits equal the average of logits obtained by applying the classification head pointwise to the feature grid prior to GAP.<n>This structure naturally suggests a multiple-instance learning (MIL) interpretation, where an image is viewed as a bag of spatial instances.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Modern image classifiers widely adopt global average pooling (GAP) followed by a linear classification head. This linearity ensures that the image-level logits equal the average of logits obtained by applying the classification head pointwise to the feature grid prior to GAP. Consequently, standard classifiers may inherently retain spatial class evidence that remains recoverable even when the image-level prediction is incorrect. This structure naturally suggests a multiple-instance learning (MIL) interpretation, where an image is viewed as a bag of spatial instances. Within this formulation, we demonstrate that standard classifiers trained with a single label per image can still learn the intended classification task in multi-object scenes. We further exploit this property to decompose image-level logits into a prediction grid, providing a post-hoc diagnostic to extract spatial class evidence that GAP otherwise obscures. Our systematic evaluation reveals that off-the-shelf models consistently recover the ground-truth class within foreground regions. The MIL interpretation further suggests that common classifier failures reflect known limitations of mean aggregation.

Related papers

• Learning to Align Generative Appearance Priors for Fine-grained Image Retrieval [54.09324791167742]
  GAPan is an alignment network that reformulates the learning objective from category prediction toward appearance modeling.<n>We show that GAPan achieves state-of-the-art performance on both widely-used fine- and coarse-grained benchmarks.
  arXiv  Detail & Related papers  (2026-05-11T01:35:26Z)
• What Are We Really Measuring? Rethinking Dataset Bias in Web-Scale Natural Image Collections via Unsupervised Semantic Clustering [11.774459974874587]
  In computer vision, a prevailing method for quantifying dataset bias is to train a model to distinguish between datasets.<n>This approach assumes that standard image augmentations successfully suppress low-level, non-semantic cues.<n>We demonstrate that this fundamental assumption is flawed within the domain of large-scale natural image collections.
  arXiv  Detail & Related papers  (2026-04-15T08:23:33Z)
• Advancing Image Classification with Discrete Diffusion Classification Modeling [25.92509040578864]
  DiDiCM is a novel framework that leverages a diffusion-based procedure to model the posterior distribution of class labels conditioned on the input image.<n>We show that a few diffusion achieve higher classification accuracy on the ImageNet dataset compared to baselines, with accuracy gains increasing as the task becomes more challenging.
  arXiv  Detail & Related papers  (2025-11-25T12:42:26Z)
• Classification non supervis{é}es d'acquisitions hyperspectrales cod{é}es : quelles v{é}rit{é}s terrain ? [4.718753827404684]
  unsupervised classification method using a limited number of coded acquisitions from a DD-CASSI hyperspectral imager.<n>Based on a simple model of intra-class spectral variability, this approach allow to identify classes and estimate reference spectra, despite data compression by a factor of ten.
  arXiv  Detail & Related papers  (2025-08-04T06:18:03Z)
• Spatial Action Unit Cues for Interpretable Deep Facial Expression Recognition [55.97779732051921]
  State-of-the-art classifiers for facial expression recognition (FER) lack interpretability, an important feature for end-users. A new learning strategy is proposed to explicitly incorporate AU cues into classifier training, allowing to train deep interpretable models. Our new strategy is generic, and can be applied to any deep CNN- or transformer-based classifier without requiring any architectural change or significant additional training time.
  arXiv  Detail & Related papers  (2024-10-01T10:42:55Z)
• Guided Interpretable Facial Expression Recognition via Spatial Action Unit Cues [55.97779732051921]
  A new learning strategy is proposed to explicitly incorporate au cues into classifier training. We show that our strategy can improve layer-wise interpretability without degrading classification performance.
  arXiv  Detail & Related papers  (2024-02-01T02:13:49Z)
• High-fidelity Pseudo-labels for Boosting Weakly-Supervised Segmentation [17.804090651425955]
  Image-level weakly-supervised segmentation (WSSS) reduces the usually vast data annotation cost by surrogate segmentation masks during training. Our work is based on two techniques for improving CAMs; importance sampling, which is a substitute for GAP, and the feature similarity loss. We reformulate both techniques based on binomial posteriors of multiple independent binary problems. This has two benefits; their performance is improved and they become more general, resulting in an add-on method that can boost virtually any WSSS method.
  arXiv  Detail & Related papers  (2023-04-05T17:43:57Z)
• Parametric Classification for Generalized Category Discovery: A Baseline Study [70.73212959385387]
  Generalized Category Discovery (GCD) aims to discover novel categories in unlabelled datasets using knowledge learned from labelled samples. We investigate the failure of parametric classifiers, verify the effectiveness of previous design choices when high-quality supervision is available, and identify unreliable pseudo-labels as a key problem. We propose a simple yet effective parametric classification method that benefits from entropy regularisation, achieves state-of-the-art performance on multiple GCD benchmarks and shows strong robustness to unknown class numbers.
  arXiv  Detail & Related papers  (2022-11-21T18:47:11Z)
• DGMIL: Distribution Guided Multiple Instance Learning for Whole Slide Image Classification [9.950131528559211]
  We propose a feature distribution guided deep MIL framework for WSI classification and positive patch localization. Experiments on the CAMELYON16 dataset and the TCGA Lung Cancer dataset show that our method achieves new SOTA for both global classification and positive patch localization tasks.
  arXiv  Detail & Related papers  (2022-06-17T16:04:30Z)
• LEAD: Self-Supervised Landmark Estimation by Aligning Distributions of Feature Similarity [49.84167231111667]
  Existing works in self-supervised landmark detection are based on learning dense (pixel-level) feature representations from an image. We introduce an approach to enhance the learning of dense equivariant representations in a self-supervised fashion. We show that having such a prior in the feature extractor helps in landmark detection, even under drastically limited number of annotations.
  arXiv  Detail & Related papers  (2022-04-06T17:48:18Z)
• Semantic Distribution-aware Contrastive Adaptation for Semantic Segmentation [50.621269117524925]
  Domain adaptive semantic segmentation refers to making predictions on a certain target domain with only annotations of a specific source domain. We present a semantic distribution-aware contrastive adaptation algorithm that enables pixel-wise representation alignment. We evaluate SDCA on multiple benchmarks, achieving considerable improvements over existing algorithms.
  arXiv  Detail & Related papers  (2021-05-11T13:21:25Z)
• Joint Visual and Temporal Consistency for Unsupervised Domain Adaptive Person Re-Identification [64.37745443119942]
  This paper jointly enforces visual and temporal consistency in the combination of a local one-hot classification and a global multi-class classification. Experimental results on three large-scale ReID datasets demonstrate the superiority of proposed method in both unsupervised and unsupervised domain adaptive ReID tasks.
  arXiv  Detail & Related papers  (2020-07-21T14:31:27Z)
• Classify and Generate: Using Classification Latent Space Representations for Image Generations [17.184760662429834]
  We propose a discriminative modeling framework that employs manipulated supervised latent representations to reconstruct and generate new samples belonging to a given class. ReGene has higher classification accuracy than existing conditional generative models while being competitive in terms of FID.
  arXiv  Detail & Related papers  (2020-04-16T09:13:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.