Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Existing deep convolutional neural networks (CNNs) require a fixed-size (e.g., 224 × 224) input image. This requirement is "artificial" and may reduce the recognition accuracy for the images or sub-images of an arbitrary size/scale. In this work, we equip the networks with another pooling strategy, "spatial pyramid pooling", to eliminate the above requirement. The new network structure, called SPP-net, can generate a fixed-length representation regardless of image size/scale. Pyramid pooling is also robust to object deformations. With these advantages, SPP-net should in general improve all CNN-based image classification methods. On the ImageNet 2012 dataset, we demonstrate that SPP-net boosts the accuracy of a variety of CNN architectures despite their different designs. On the Pascal VOC 2007 and Caltech101 datasets, SPP-net achieves state-of-the-art classification results using a single full-image representation and no fine-tuning. The power of SPP-net is also significant in object detection. Using SPP-net, we compute the feature maps from the entire image only once, and then pool features in arbitrary regions (sub-images) to generate fixed-length representations for training the detectors. This method avoids repeatedly computing the convolutional features. In processing test images, our method is 24-102 × faster than the R-CNN method, while achieving better or comparable accuracy on Pascal VOC 2007. In ImageNet Large Scale Visual Recognition Challenge (ILSVRC) 2014, our methods rank #2 in object detection and #3 in image classification among all 38 teams. This manuscript also introduces the improvement made for this competition.

Last synced 2026-08-22 14:19 UTC

• SFC-DETR: a small object detection model for drone images based on spatial and frequency domain collaborative enhancement

  PeerJ Computer Science 2026 · 0 citations

• A Novel Weight Initialization Scheme for Randomized Leaky Rectified Linear Units and a Comparative Analysis of Rectified Activation Functions for Image Classification

  Electronics 2026 · 0 citations

• AC-UKAN: An enhanced U-KAN framework integrating modulated deformable convolution and mixed pooling

  Biomedical Signal Processing and Control 2026 · 0 citations

• GMFC-YOLO11n: a lightweight object detection network for microscopic goose and duck down identification

  Measurement Science and Technology 2026 · 0 citations

• Multiscale attention guided historical prompt tracking for industrial particles

  Discover Computing 2026 · 0 citations

• Compact Visual Detection of Small Drones Through Contextual Statistics and Geometric Consistency

  Applied Sciences 2026 · 0 citations

• Wavelet Interactive Mamba Fusion Network for UAV Small Object Detection with Visible–Thermal Data

  Sensors 2026 · 0 citations

• From YOLOv1 to YOLO26: A Unified Comparative Review of the Evolution, Architecture, and Applications of YOLO-Based Object Detection Frameworks

  Applied Sciences 2026 · 0 citations

• Spatial reconstruction feature for object detection on coal mine

  Frontiers in Artificial Intelligence 2026 · 0 citations