mlsanityData to 7 Oct 2026

Quantization-Guided Training for Compact TinyML Models

arXiv 2103.06231 ↗First posted 10 Mar 2021cs.LG cs.CVcited by 10

Not on the map. Why?

Abstract (arXiv)

We propose a Quantization Guided Training (QGT) method to guide DNN training towards optimized low-bit-precision targets and reach extreme compression levels below 8-bit precision. Unlike standard quantization-aware training (QAT) approaches, QGT uses customized regularization to encourage weight values towards a distribution that maximizes accuracy while reducing quantization errors. One of the main benefits of this approach is the ability to identify compression bottlenecks. We validate QGT using state-of-the-art model architectures on vision datasets. We also demonstrate the effectiveness of QGT with an 81KB tiny model for person detection down to 2-bit precision (representing 17.7x size reduction), while maintaining an accuracy drop of only 3% compared to a floating-point baseline.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 23 of its 37 references that are in the map. First posted on arXiv 10 Mar 2021.

Builds on

23 of 37 references.

References · 23

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  3. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  4. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  5. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,773
  6. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,291
  7. 2018The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networkscited by 2,250
  8. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,548
  9. 2017ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devicescited by 1,194
  10. 2016Pruning Filters for Efficient ConvNetscited by 1,129
  11. 2019Once-for-All: Train One Network and Specialize it for Efficient Deploymentcited by 778
  12. 2017Learning Sparse Neural Networks through L0 Regularizationcited by 571
11 more
  1. 2017A systematic study of the class imbalance problem in convolutional neural networkscited by 510
  2. 2018Quantizing deep convolutional networks for efficient inference: A whitepapercited by 507
  3. 2020Training with Quantization Noise for Extreme Model Compressioncited by 145
  4. 2019Well-Read Students Learn Better: On the Importance of Pre-training Compact Modelscited by 118
  5. 2020Train Large, Then Compress: Rethinking Model Size for Efficient Training and Inference of Transformerscited by 67
  6. 2019Visual Wake Words Datasetcited by 46
  7. 2018Discovering Low-Precision Networks Close to Full-Precision Networks for Efficient Embedded Inferencecited by 33
  8. 2019Bit Efficient Quantization for Deep Neural Networkscited by 13
  9. 2018Generalized Ternary Connect: End-to-End Learning and Compression of Multiplication-Free Deep Neural Networkscited by 3
  10. 2019SinReQ: Generalized Sinusoidal Regularization for Low-Bitwidth Deep Quantized Trainingcited by 2
  11. 2020Subtensor Quantization for Mobilenetscited by 1

Built on by

Cited by year

Citing papers per year: 2021: 2; 2022: 4; 2023: 0; 2024: 2; 2025: 2.

10 papers.

  1. 2021Implicit Neural Representations for Image Compressioncited by 161
  2. 2022Overcoming Oscillations in Quantization-Aware Trainingcited by 116
  3. 2021A TinyML Platform for On-Device Continual Learning with Quantized Latent Replayscited by 37
  4. 2022An Empirical Study of Low Precision Quantization for TinyMLcited by 6
  5. 2025From Tiny Machine Learning to Tiny Deep Learning: A Surveycited by 5
  6. 2025Quantizing Small-Scale State-Space Models for Edge AIcited by 3
  7. 2022LightDepth: A Resource Efficient Depth Estimation Approach for Dealing with Ground Truth Sparsity via Curriculum Learningcited by 2
  8. 2024Extreme Compression of Adaptive Neural Imagescited by 1
  9. 2022Deep learning model compression using network sensitivity and gradientscited by 0
  10. 2024Toward Efficient Convolutional Neural Networks With Structured Ternary Patternscited by 0