[1710.03958] Detect to Track and Track to Detect

Recent approaches for high accuracy detection and tracking of object categories in video consist of complex multistage solutions that become more cumbersome each year. In this paper we propose a ConvNet architecture that jointly performs detection and tracking, solving the task in a simple and effective way. Our contributions are threefold: (i) we set up a ConvNet architecture for simultaneous detection and tracking, using a multi-task objective for frame-based object detection and across-frame track regression; (ii) we introduce correlation features that represent object co-occurrences across time to aid the ConvNet during tracking; and (iii) we link the frame level detections based on our across-frame tracklets to produce high accuracy detections at the video level. Our ConvNet architecture for spatiotemporal object detection is evaluated on the large-scale ImageNet VID dataset where it achieves state-of-the-art results. Our approach provides better single model performance than the

1 mentions: @Maxwell_110
Date:

Referring Tweets

@Maxwell_110
@Maxwell_110 Detect to Track 📝 t.co/xtmuDFhZ5H Detection と Tracking を同時に行う R-FCN(t.co/vQUhS1zV46)base のビデオ物体検出モデルを提案 bbox のフレーム間変化量(x, y, h, w)を MTLoss で学習するが(Eq. 1),フレーム feature map 間の相関を RoI-Pooling に追加入力して学習 t.co/Ei2Vl3MYYO

Related Entries

Deep Learning 1: Introduction to Machine Learning Based AI - YouTube
Read more Deep Learning 1: Introduction to Machine Learning Based AI - YouTube
0 users, 1 mentions 2019/03/24 12:47
RSNA-MICCAI Brain Tumor Radiogenomic Classification | Kaggle
Read more RSNA-MICCAI Brain Tumor Radiogenomic Classification | Kaggle
0 users, 1 mentions 2021/07/14 00:12
AIcrowd | Getting Started | Posts
Read more AIcrowd | Getting Started | Posts
0 users, 1 mentions 2021/07/31 04:38
[2109.14172] Semi-Supervised Segmentation of Radiation-Induced Pulmonary Fibrosis from Lung CT Scans...
Read more [2109.14172] Semi-Supervised Segmentation of Radiation-Induced Pulmonary Fibrosis from Lung CT Scans...
0 users, 1 mentions 2021/10/28 22:37
[2203.02378v1] DiT: Self-supervised Pre-training for Document Image Transformer
Read more [2203.02378v1] DiT: Self-supervised Pre-training for Document Image Transformer
0 users, 1 mentions 2022/04/05 22:37

ML-Newsについて

機械学習の技術に関する情報は流速も早いし、分野も多様でキャッチアップが大変です。Twitterで機械学習用のリストを作っても、普段は機械学習以外の話題が多く流れており、効率的に情報収集するのは困難です。

ML-NewsはSNSを情報源とした機械学習に特化したニュースサイトです。機械学習に関する論文、ブログ、ライブラリ、コンペティション、発表資料、勉強会などの最新の情報を効率的に収集できます。

機械学習を応用した自然言語処理、画像認識、情報検索などの分野の情報や機械学習で必要になるデータ基盤やMLOpsの話題もカバーしています。
安定したサイト運営のためにGitHub sponsorを募集しています。

お知らせ

  • 2021/12/31: デザインを刷新しました
  • 2021/04/08: 日本語とKaggleのカテゴリを新設しました