mlsanityData to 7 Oct 2026

On-Sensor Binarized Fully Convolutional Neural Network with A Pixel Processor Array

arXiv 2202.00836 ↗First posted 2 Feb 2022cs.CV cs.ET eess.SPcited by 0

Not on the map. Why?

Abstract (arXiv)

This work presents a method to implement fully convolutional neural networks (FCNs) on Pixel Processor Array (PPA) sensors, and demonstrates coarse segmentation and object localisation tasks. We design and train binarized FCN for both binary weights and activations using batchnorm, group convolution, and learnable threshold for binarization, producing networks small enough to be embedded on the focal plane of the PPA, with limited local memory resources, and using parallel elementary add/subtract, shifting, and bit operations only. We demonstrate the first implementation of an FCN on a PPA device, performing three convolution layers entirely in the pixel-level processors. We use this architecture to demonstrate inference generating heat maps for object segmentation and localisation at over 280 FPS using the SCAMP-5 PPA vision chip.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 16 of its 32 references that are in the map. First posted on arXiv 2 Feb 2022.

Builds on

16 of 32 references.

References · 16

  1. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  2. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,845
  3. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
  4. 2015BinaryConnect: Training Deep Neural Networks with binary weights during propagationscited by 571
  5. 2016R-FCN: Object Detection via Region-based Fully Convolutional Networkscited by 414
  6. 2020ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functionscited by 183
  7. 2021TEyeD: Over 20 million real-world eye images with Pupil, Eyelid, and Iris 2D and 3D Segmentations, 2D and 3D Landmarks, 3D Eyeball, Gaze Vector, and Eye Movement Typescited by 32
  8. 2018TernausNetV2: Fully Convolutional Network for Instance Segmentationcited by 25
  9. 2019Fully Learnable Group Convolution for Acceleration of Deep Neural Networkscited by 16
  10. 2020Fully Embedding Fast Convolutional Networks on Pixel Processor Arrayscited by 13
  11. 2019A Camera That CNNs: Towards Embedded Neural Networks on Pixel Processor Arrayscited by 9
  12. 2019How Does Batch Normalization Help Binary Training?cited by 7
4 more
  1. 2020AnalogNet: Convolutional Neural Network Inference on Analog Focal Plane Sensor Processorscited by 4
  2. 2020Agile Reactive Navigation for A Non-Holonomic Mobile Robot Using A Pixel Processor Arraycited by 1
  3. 2021Bringing A Robot Simulator to the SCAMP Vision Systemcited by 1
  4. 2021Direct Servo Control from In-Sensor CNN Inference with A Pixel Processor Arraycited by 1

Built on by

Not cited yet.