mlsanityData to 7 Oct 2026

Improved Projection Learning for Lower Dimensional Feature Maps

arXiv 2210.15170 ↗First posted 27 Oct 2022cs.LG cs.CVcited by 2

Not on the map. Why?

Abstract (arXiv)

The requirement to repeatedly move large feature maps off- and on-chip during inference with convolutional neural networks (CNNs) imposes high costs in terms of both energy and time. In this work we explore an improved method for compressing all feature maps of pre-trained CNNs to below a specified limit. This is done by means of learned projections trained via end-to-end finetuning, which can then be folded and fused into the pre-trained network. We also introduce a new `ceiling compression' framework in which evaluate such techniques in view of the future goal of performing inference fully on-chip.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 14 of its 19 references that are in the map. First posted on arXiv 27 Oct 2022.

Builds on

14 of 19 references.

References · 14

  1. 2021A Survey of Quantization Methods for Efficient Neural Network Inferencecited by 798
  2. 2020What is the State of Neural Network Pruning?cited by 578
  3. 2017Efficient Processing of Deep Neural Networks: A Tutorial and Surveycited by 427
  4. 2016EIE: Efficient Inference Engine on Compressed Deep Neural Networkcited by 291
  5. 2017SCNN: An Accelerator for Compressed-sparse Convolutional Neural Networkscited by 132
  6. 2017NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Mapscited by 39
  7. 2017Compressing DMA Engine: Leveraging Activation Sparsity for Training Deep Neural Networkscited by 36
  8. 2018Accelerating Convolutional Neural Networks via Activation Map Compressioncited by 34
  9. 2014Efficient and Accurate Approximations of Nonlinear Convolutional Networkscited by 30
  10. 2019Feature Map Transform Coding for Energy-Efficient CNN Inferencecited by 10
  11. 2019EBPC: Extended Bit-Plane Compression for Deep Neural Network Inference and Training Acceleratorscited by 4
  12. 2021Memory-Efficient CNN Accelerator Based on Interlayer Feature Map Compressioncited by 4
2 more
  1. 2021Transform-Based Feature Map Compression for CNN Inferencecited by 3
  2. 2021Compression-aware Projection with Greedy Dimension Reduction for Convolutional Neural Network Activationscited by 3

Built on by

Cited by year

Citing papers per year: 2023: 1; 2024: 0; 2025: 1.

2 papers.

  1. 2023On the Initialisation of Wide Low-Rank Feedforward Neural Networkscited by 5
  2. 2025Theoretical Guarantees for Low-Rank Compression of Deep Neural Networkscited by 3