mlsanityData to 14 Oct 2026

Fully Quantized Always-on Face Detector Considering Mobile Image Sensors

arXiv 2311.01001 ↗First posted 2 Nov 2023cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Despite significant research on lightweight deep neural networks (DNNs) designed for edge devices, the current face detectors do not fully meet the requirements for "intelligent" CMOS image sensors (iCISs) integrated with embedded DNNs. These sensors are essential in various practical applications, such as energy-efficient mobile phones and surveillance systems with always-on capabilities. One noteworthy limitation is the absence of suitable face detectors for the always-on scenario, a crucial aspect of image sensor-level applications. These detectors must operate directly with sensor RAW data before the image signal processor (ISP) takes over. This gap poses a significant challenge in achieving optimal performance in such scenarios. Further research and development are necessary to bridge this gap and fully leverage the potential of iCIS applications. In this study, we aim to bridge the gap by exploring extremely low-bit lightweight face detectors, focusing on the always-on face detection scenario for mobile image sensor applications. To achieve this, our proposed model utilizes sensor-aware synthetic RAW inputs, simulating always-on face detection processed "before" the ISP chain. Our approach employs ternary (-1, 0, 1) weights for potential implementations in image sensors, resulting in a relatively simple network architecture with shallow layers and extremely low-bitwidth. Our method demonstrates reasonable face detection performance and excellent efficiency in simulation studies, offering promising possibilities for practical always-on face detectors in real-world applications.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 44 of its 81 references that are in the map. First posted on arXiv 2 Nov 2023.

Builds on

44 of 81 references.

References · 44

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  4. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  5. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  6. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,845
  7. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  8. 2013Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computationcited by 2,079
  9. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,593
  10. 2020Sharpness-Aware Minimization for Efficiently Improving Generalizationcited by 1,417
  11. 2016SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model sizecited by 1,060
  12. 2021A Survey of Quantization Methods for Efficient Neural Network Inferencecited by 815
32 more
  1. 2019Learned Step Size Quantizationcited by 628
  2. 2021A White Paper on Neural Network Quantizationcited by 472
  3. 2018PACT: Parameterized Clipping Activation for Quantized Neural Networkscited by 467
  4. 2015WIDER FACE: A Face Detection Benchmarkcited by 298
  5. 2018LQ-Nets: Learned Quantization for Highly Accurate and Compact Deep Neural Networkscited by 274
  6. 2018Unprocessing Images for Learned Raw Denoisingcited by 250
  7. 2018MobileFaceNets: Efficient CNNs for Accurate Real-Time Face Verification on Mobile Devicescited by 182
  8. 2019Quantization Networkscited by 153
  9. 2018DSFD: Dual Shot Face Detectorcited by 116
  10. 2019Zoom To Learn, Learn To Zoomcited by 115
  11. 2019BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUscited by 74
  12. 2020PROFIT: A Novel Training Method for sub-4-bit MobileNet Modelscited by 50
  13. 2022BokehMe: When Neural Rendering Meets Classical Renderingcited by 48
  14. 2017FaceBoxes: A CPU Real-time Face Detector with High Accuracycited by 45
  15. 2020Searching for Low-Bit Weights in Quantized Neural Networkscited by 44
  16. 2019Training Quantized Neural Networks with a Full-precision Auxiliary Modulecited by 35
  17. 2022NIPQ: Noise proxy-based Integrated Pseudo-Quantizationcited by 35
  18. 2021Learned Smartphone ISP on Mobile NPUs with Deep Learning, Mobile AI 2021 Challenge: Reportcited by 34
  19. 2018Discovering Low-Precision Networks Close to Full-Precision Networks for Efficient Embedded Inferencecited by 33
  20. 2022CADyQ: Content-Aware Dynamic Quantization for Image Super-Resolutioncited by 27
  21. 2018NICE: Noise Injection and Clamping Estimation for Neural Network Quantizationcited by 24
  22. 2019RefineFace: Refinement Neural Network for High Performance Face Detectioncited by 21
  23. 2020Fully Quantized Image Super-Resolution Networkscited by 19
  24. 2022QuantFace: Towards Lightweight Face Recognition by Synthetic Data Low-bit Quantizationcited by 18
  25. 2019AirFace: Lightweight and Efficient Model for Face Recognitioncited by 9
  26. 2017LCDet: Low-Complexity Fully-Convolutional Neural Networks for Object Detection in Embedded Systemscited by 7
  27. 2021Detect Faces Efficiently: A Survey and Evaluationscited by 6
  28. 2019Face Detection with Feature Pyramids and Landmarkscited by 5
  29. 2020Real-time Mask Detection on Google Edge TPUcited by 3
  30. 2019IFQ-Net: Integrated Fixed-point Quantization Networks for Embedded Visioncited by 2
  31. 2022Real-time Hyper-Dimensional Reconfiguration at the Edge using Hardware Acceleratorscited by 2
  32. 2019DupNet: Towards Very Tiny Quantized CNN with Improved Accuracy for Face Detectioncited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2022Adaptation of MobileNetV2 for Face Detection on Ultra-Low Power Platform
  2. 2022LPYOLO: Low Precision YOLO for Face Detection on FPGA
  3. 2023Flexible and Fully Quantized Ultra-Lightweight TinyissimoYOLO for Ultra-Low-Power Edge Systems
  4. 2025J3DAI: A tiny DNN-Based Edge AI Accelerator for 3D-Stacked CMOS Image Sensor
  5. 2025On-Sensor Convolutional Neural Networks with Early-Exits
  6. 2023TinyissimoYOLO: A Quantized, Low-Memory Footprint, TinyML Object Detection Network for Low Power Microcontrollers
  7. 2026Enabling Fully Integer-Only Inference for Lightweight Detection Transformers
  8. 2025Designing Object Detection Models for TinyML: Foundations, Comparative Analysis, Challenges, and Emerging Solutions