WeChat Mini Program
Old Version Features

Research on Instance-level Video Object Detection Based on Momentum

2023 International Conference on Intelligent Media, Big Data and Knowledge Mining (IMBDKM)(2023)

School of Computer Science and Engineering

Cited 0|Views8
Abstract
Since the targets in video have dynamic features, such as appearance, shape, and size, etc. Therefore, the method of frame by frame detection of static images is directly applied to video target detection, which is difficult to meet the requirements of video target detection for spatio- temporal consistency. The traditional solution is to aggregate the non-key frame features near the key frames, enhance each key frame, and finally transfer the enhanced key frame feature map to the static image target detection method, but such methods lack the utilization of the time-domain information between video frames, resulting in low detection accuracy. To address these shortcomings, this paper proposes a momentum-based instance-level video target detection method to improve the detection accuracy of the model. The method fully considers the time-domain relationship between the instance-level calibration corresponding to each key frame in the time-domain, obtains the video momentum-level calibration features by bonding the pixel-level features and the instance-level calibration features, and fuses the pixel-level calibration with the momentum-level calibration to enhance the model robustness. In this paper, we experimentally validate the ImageNet VID dataset, and the results show the effectiveness of the momentum-based instance-level video target detection algorithm with an mAP accuracy of 81.2%.
More
Translated text
Key words
video object detection,time domain information,pixel-level calibration,instance-level calibration,momentum- level calibration
求助PDF
上传PDF
Bibtex
AI Read Science
AI Summary
AI Summary is the key point extracted automatically understanding the full text of the paper, including the background, methods, results, conclusions, icons and other key content, so that you can get the outline of the paper at a glance.
Example
Background
Key content
Introduction
Methods
Results
Related work
Fund
Key content
  • Pretraining has recently greatly promoted the development of natural language processing (NLP)
  • We show that M6 outperforms the baselines in multimodal downstream tasks, and the large M6 with 10 parameters can reach a better performance
  • We propose a method called M6 that is able to process information of multiple modalities and perform both single-modal and cross-modal understanding and generation
  • The model is scaled to large model with 10 billion parameters with sophisticated deployment, and the 10 -parameter M6-large is the largest pretrained model in Chinese
  • Experimental results show that our proposed M6 outperforms the baseline in a number of downstream tasks concerning both single modality and multiple modalities We will continue the pretraining of extremely large models by increasing data to explore the limit of its performance
Upload PDF to Generate Summary
Must-Reading Tree
Example
Generate MRT to find the research sequence of this paper
Data Disclaimer
The page data are from open Internet sources, cooperative publishers and automatic analysis results through AI technology. We do not make any commitments and guarantees for the validity, accuracy, correctness, reliability, completeness and timeliness of the page data. If you have any questions, please contact us by email: report@aminer.cn
Chat Paper
Summary is being generated by the instructions you defined