AdaPool: Exponential Adaptive Pooling for Information-Retaining Downsampling

Alexandros Stergiou, Ronald Poppe

Research output: Contribution to journalArticleAcademicpeer-review

2 Downloads (Pure)

Abstract

Pooling layers are essential building blocks of convolutional neural networks (CNNs), to reduce computational overhead and increase the receptive fields of proceeding convolutional operations. Their goal is to produce downsampled volumes that closely resemble the input volume while, ideally, also being computationally and memory efficient. Meeting both these requirements remains a challenge. To this end, we propose an adaptive and exponentially weighted pooling method: adaPool. Our method learns a regional-specific fusion of two sets of pooling kernels that are based on the exponent of the Dice-Sørensen coefficient and the exponential maximum, respectively. AdaPool improves the preservation of detail on a range of tasks including image and video classification and object detection. A key property of adaPool is its bidirectional nature. In contrast to common pooling methods, the learned weights can also be used to upsample activation maps. We term this method adaUnPool. We evaluate adaUnPool on image and video super-resolution and frame interpolation. For benchmarking, we introduce Inter4K, a novel high-quality, high frame-rate video dataset. Our experiments demonstrate that adaPool systematically achieves better results across tasks and backbones, while introducing a minor additional computational and memory overhead.
Original languageEnglish
Pages (from-to)251-266
Number of pages16
JournalIEEE transactions on image processing
Volume32
Early online date12 Dec 2022
DOIs
Publication statusPublished - 2023
Externally publishedYes

Keywords

  • Computer architecture
  • Interpolation
  • Kernel
  • Pooling
  • Superresolution
  • Task analysis
  • Visualization
  • Weight measurement
  • Downsampling
  • Upsampling

Fingerprint

Dive into the research topics of 'AdaPool: Exponential Adaptive Pooling for Information-Retaining Downsampling'. Together they form a unique fingerprint.

Cite this