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 Concept Videos

Deconvolution01:20

Deconvolution

377
Deconvolution, also known as inverse filtering, is the process of extracting the impulse response from known input and output signals. This technique is vital in scenarios where the system's characteristics are unknown, and they must be inferred from the observable signals.
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...
377
Convolution Properties II01:17

Convolution Properties II

359
The important convolution properties include width, area, differentiation, and integration properties.
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
359
Convolution Properties I01:20

Convolution Properties I

323
Convolution computations can be simplified by utilizing their inherent properties.
The commutative property reveals that the input and the impulse response of an LTI (Linear Time-Invariant) system can be interchanged without affecting the output:
323
Difference from Background: Limit of Detection01:05

Difference from Background: Limit of Detection

7.5K
The limit of detection (LOD) is the smallest amount of analyte that can be distinguished from the background noise. The LOD value corresponds to the concentration at which the analyte signal is three times larger than the standard deviation of the blank signal. Below this value, the analyte signal cannot be differentiated from the background noise. It is calculated by dividing the calibration slope by 3 times the standard deviation of the blank signals.
The LOD indicates the presence or absence...
7.5K
Convolution: Math, Graphics, and Discrete Signals01:24

Convolution: Math, Graphics, and Discrete Signals

559
In any LTI (Linear Time-Invariant) system, the convolution of two signals is denoted using a convolution operator, assuming all initial conditions are zero. The convolution integral can be divided into two parts: the zero-input or natural response and the zero-state or forced response, with t0 indicating the initial time.
To simplify the convolution integral, it is assumed that both the input signal and impulse response are zero for negative time values. The graphical convolution process...
559

You might also read

Related Articles

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

Sort by
Same author

Epigenetic age acceleration mediates the association between pro-inflammatory and pro-oxidant diets and the progression and mortality of cardiovascular-kidney-metabolic syndrome.

Genes & nutrition·2026
Same author

A high-strength radiative cooling biocomposite via hierarchical crystalline nanostructuring.

Nature communications·2026
Same author

Dynamic variability of serum sodium, potassium, and calcium and mortality after acute myocardial infarction: insights from traditional and machine learning approaches.

BMC cardiovascular disorders·2026
Same author

The cholesterol, high-density lipoprotein, and glucose index and its derived indices predict diabetic kidney disease development and progression: evidence from two nationwide cohorts and a biopsy-proven clinical cohort.

Cardiovascular diabetology·2026
Same author

Joint impact of stress hyperglycaemic ratio and glycaemic variability in patients with ischaemic stroke and machine learning for mortality prediction.

BMC neurology·2025
Same author

COL6A2: A Key Survival-Related Gene and Restricting Antitumor Immunity in Glioblastoma.

Cancer science·2025

Related Experiment Video

Updated: Nov 8, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

749

A new multi-scale backbone network for object detection based on asymmetric convolutions.

Xianghua Ma1, Zhenkun Yang1

  • 1School of Electrical and Electronic Engineering, Shanghai Institute of Technology, Shanghai, China.

Science Progress
|April 21, 2021
PubMed
Summary

This study introduces a new Multi-scale Asymmetric Convolution (MAC) block to improve object detection accuracy on mobile devices. The MAC block enhances feature extraction, boosting precision without sacrificing speed.

Keywords:
Object detectionasymmetric convolutionsbackbone networkdeep learningmulti-scale representation

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

620
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.1K

Related Experiment Videos

Last Updated: Nov 8, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

749
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

620
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.1K

Area of Science:

  • Computer Vision
  • Deep Learning

Background:

  • Real-time object detection on mobile platforms faces a trade-off between speed and accuracy.
  • Lightweight object detectors are fast but lack precision.
  • Extracting complete multi-scale image features is key to improving detection accuracy.

Purpose of the Study:

  • To propose a novel Multi-scale Asymmetric Convolution (MAC) block to enhance the multi-scale representation ability of Convolutional Neural Networks (CNNs).
  • To improve object detection accuracy on mobile platforms by enhancing feature extraction.

Main Methods:

  • Exploited three different asymmetric convolutions in parallel to create the MAC block.
  • Developed a method for MAC blocks to adaptively merge features of different scales using learnable weighted parameters.
  • Integrated MAC blocks into state-of-the-art backbones like ResNet-50 to form new multi-scale backbone networks.

Main Results:

  • Experimental results on CIFAR-100, PASCAL VOC, and MS COCO datasets demonstrated significant improvements in detection precision.
  • The proposed MAC block maintained a fast detection speed, addressing the speed-accuracy trade-off.
  • The MAC block effectively enhanced the multi-scale representation capabilities of CNNs.

Conclusions:

  • The MAC block is an effective module for improving object detection accuracy, particularly for multi-scale objects.
  • The proposed approach offers a viable solution for achieving high-precision, real-time object detection on mobile devices.
  • MAC blocks can be seamlessly integrated into existing object detection frameworks to boost performance.