Abstract:Feature pooling is a key component in modern visual classification system. However, the conventional two prevailing pooling techniques, namely average and max poolings, are not theoretically optimal, due to the unrecoverable loss of the spatial information during the statistical summarization and the underlying over-simplified assumption about the feature distribution. Addressing these issues, this paper proposes to generalize previous pooling methods toward a weighted ℓp-norm spatial pooling function tailored for class-specific feature spatial distribution. Optimizing such a pooling function toward discriminative class separability that is subject to a spatial smoothness constraint yields a so-called geometric ℓp-norm pooling (GLP) method. Furthermore, to handle the variation of object scale/position, which would affect not only the learning of discriminative pooling weights but also the applicability of the learned weights, we propose a simple yet effective self-alignment step during both learning and testing to adaptively adjust the pooling weights for individual images. Image segmentation and visual saliency map are utilized to construct a directed pixel adjacency graph. The discriminative pooling weights are diffused using random walk on the constructed graph and therefore the discriminative pooling weights are propagated onto the salient and foreground region. This leads to a robust version of GLP (RGLP) which can cope with the misalignment of object position and scale in images. Comprehensive experiments validate the effectiveness of the proposed GLP feature pooling framework. The proposed random walk based self-alignment step can effectively alleviate the image misalignment issue and further boost classification accuracy. State-of-the-art image classification and action recognition performances are attained on several benchmarks.

Gradient Corner Pooling for Keypoint-Based Object Detection

PoolNet plus : Exploring the Potential of Pooling for Salient Object Detection

CenterNet: Keypoint Triplets for Object Detection

CornerNet: Detecting Objects as Paired Keypoints

Key Point Detection By Max Pooling For Tracking

A Simple Pooling-Based Design for Real-Time Salient Object Detection.

Corner2Net: Detecting Objects as Cascade Corners

Object Detection Based on Faster R-CNN Algorithm with Skip Pooling and Fusion of Contextual Information

Corner Proposal Network for Anchor-Free, Two-Stage Object Detection

Lightweight Anchor-free Object Detection Algorithm Based on Keypoint Detection

An anchor-free object detector with novel corner matching method

A Pooling-Based Feature Pyramid Network for Salient Object Detection

Salient Object Detection Network with Center Pooling and Distance-Weighted Affinity Loss Function

Object recognition method by new vision-based pooling operation

Pyramid Sub-Region Sensitive Network for Object Detection.

Enhanced Pooling Method for Convolutional Neural Networks Based on Optimal Search Theory

Pooling Pyramid Network for Object Detection

Robust Geometric ℓp-Norm Feature Pooling for Image Classification and Action Recognition

Pedestrian Detection Based on Adaptive Pooling Method

Mixed Pooling For Convolutional Neural Networks