Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

EQISP: efficient quantized image signal processing with multi-scale pyramid fusion for resource constrained embodied

Tongxin Yang1, Ling Guo2, Jie Li1,3

  • 1School of Naval Architecture, Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai, China.

Frontiers in Neurorobotics
|June 24, 2026
PubMed
Summary

Related Concept Videos

Parallel Processing01:20

Parallel Processing

The brain processes sensory information rapidly due to parallel processing, which involves sending data across multiple neural pathways at the same time. This method allows the brain to manage various sensory qualities, such as shapes, colors, movements, and locations, all concurrently. For instance, when observing a forest landscape, the brain simultaneously processes the movement of leaves, the shapes of trees, the depth between them, and the various shades of green. This enables a quick and...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Further improvement in London's air quality demands more than the Ultra Low Emission Zone policy.

NPJ clean air·2025
Same author

TDP-43: unveiling the hidden key to cellular fate decisions.

Cell communication and signaling : CCS·2025
Same author

Structurally Confined Ni with Oxygen-Deficient CeO<sub>2</sub> for Efficient Self-Transfer Hydrogenolysis of Lignin into Jet Fuel Precursors.

Small (Weinheim an der Bergstrasse, Germany)·2025
Same author

A Nanolasing-Based Sensor for Ultra-Sensitive Detection of Trace HSA in Artificial Urine.

Small (Weinheim an der Bergstrasse, Germany)·2025
Same author

Impact of neoadjuvant and adjuvant chemotherapy on breast cancer prognosis in a propensity score matched population.

Scientific reports·2025
Same author

A comparative evaluation of multiple machine learning approaches for forecasting dengue outbreaks in Bangladesh.

Scientific reports·2025
This summary is machine-generated.

This study introduces the enhanced quantized image signal processor (EQISP) for resource-constrained robots. EQISP significantly reduces computational load and memory usage while maintaining high image fidelity for efficient visual perception.

Area of Science:

  • Computer Vision
  • Robotics
  • Image Processing

Background:

  • Autonomous robots require efficient visual perception in complex environments.
  • Processing high dynamic range RAW sensor data is computationally expensive for edge devices.

Purpose of the Study:

  • To develop an efficient and scalable visual front-end for resource-constrained embodied perception systems.
  • To reduce the computational cost and memory footprint of neural image signal processors.

Main Methods:

  • Proposed the enhanced quantized image signal processor (EQISP), integrating a quantized convolutional neural network (QCNN) with dynamic fixed-point hybrid quantization.
  • Implemented a unified pyramid fusion algorithm (UPFA) using Gaussian and Laplacian pyramids for multi-scale, multi-exposure fusion and iterative reconstruction.
Keywords:
AI-ISPedge visionembodied intelligent systemsembodied perceptionpyramid fusionquantization

Related Experiment Videos

Main Results:

  • EQISP achieved a PSNR of 22.90 dB and SSIM of 0.9278, with 164.843 GFLOPs.
  • Demonstrated a 1.71 dB PSNR improvement and a 4.24x reduction in computational cost compared to the PyNET baseline.
  • On an NVIDIA Jetson TX2, EQISP achieved a 57 MB model size, 189 ms latency, 6.1 FPS speed, and 2.2 GB peak memory usage.

Conclusions:

  • EQISP offers a practical solution for efficient visual processing in resource-constrained embodied systems.
  • The proposed methods effectively mitigate detail loss induced by quantization while reducing computational demands.