Faster Neural Networks Straight from JPEG

By Lionel Gueguen, Alexander Sergeev, B. Kadlec, Rosanne Liu, Jason Yosinski. 2018

259 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

MgSvF: Multi-Grained Slow versus Fast Framework for Few-Shot Class-Incremental Learning (2020, background): “Some of these frequency-aware methods aim to reduce the computing cost with frequency transformation [83], [84], thus improving the network efﬁ-ciency.”
Checkmate: Breaking the Memory Wall with Optimal Tensor Rematerialization (2019, background): “Gueguen et al. [24] develop a DNN that classiﬁes ImageNet images given discrete cosine transform codes from partially decoded JPEG images, accelerating network evaluation, and possibly requiring less memory.”
Non-Discriminative Data or Weak Model? On the Relative Importance of Data and Model Resolution (2019, background): “Interestingly, in the recent work [19] also observe the maze-like structure of the first layers in architectures that try to learn DCT-like transform and use relatively large 8× 8 filters.”
Online Learned Continual Compression with Stacked Quantization Module (2019, background): “In general, compression techniques can greatly improve data storage capacity, and, if done well, reduce the memory and compute usage in downstream machine learning tasks (Gueguen et al., 2018; Oyallon et al., 2018).”
FcaNet: Frequency Channel Attention Networks (2020, methodology): “In [8, 13], frequency analysis is introduced in the CNNs by JPEG encoding.”
Online Learned Continual Compression with Adaptive Quantization Modules (2019, background): “In general, compression techniques can greatly improve data storage capacity, and, if done well, reduce the memory and compute usage in downstream machine learning tasks (Gueguen et al., 2018; Oyallon et al., 2018).”
Focal Frequency Loss for Image Reconstruction and Synthesis (2020, background): “Besides, several studies have incorporated frequency analysis with network compression [16, 66, 68, 6, 17] and feature reduction [35, 61] to accelerate the training and inference of networks.”
Photorealistic Style Transfer via Wavelet Transforms (2019, background): “There have been several attempts to incorporate both classical signal processing and deep learning approaches, including feature reduction [16], network compression [11, 16], superresolution [2], classification [3, 7, 23, 27, 30] and image denoising [14].”
Variable-rate discrete representation learning (2021, background): “Some models operate directly on compressed representations to achieve a similar effect (Nash et al., 2021), though usually some topological structure is preserved (Gueguen et al., 2018; Wu et al., 2018a).”
Frequency-aware Discriminative Feature Learning Supervised by Single-Center Loss for Face Forgery Detection (2021, background): “In order to tackle this problem, inspired by [16, 44, 31, 26], we pro-pose an adaptive frequency feature generation module (AF-FGM) to efﬁciently mine subtle artifacts from the frequency domain.”