[2112.02250] Dense Extreme Inception Network for Edge Detection

Edge detection is the basis of many computer vision applications. State of the art predominantly relies on deep learning with two decisive factors: dataset content and network's architecture. Most of the publicly available datasets are not curated for edge detection tasks. Here, we offer a solution to this constraint. First, we argue that edges, contours and boundaries, despite their overlaps, are three distinct visual features requiring separate benchmark datasets. To this end, we present a new dataset of edges. Second, we propose a novel architecture, termed Dense Extreme Inception Network for Edge Detection (DexiNed), that can be trained from scratch without any pre-trained weights. DexiNed outperforms other algorithms in the presented dataset. It also generalizes well to other datasets without any fine-tuning. The higher quality of DexiNed is also perceptually evident thanks to the sharper and finer edges it outputs.

1 mentions: @allowfirm
Date:

Referring Tweets

@allowfirm
@allowfirm CNNベースでエッジを検出する「DexiNed」のarXiv論文( t.co/f4C5rudty9 )とPyTorch実装( t.co/DHtRANSh3R )が、いずれも「内容が整理」されていて役立ちそうと思いました。タスクやデータセットを再確認してから、わりとシンプルに解くアプローチが好印象です。

Related Entries

[1910.05448] Neural Memory Plasticity for Anomaly Detection
Read more [1910.05448] Neural Memory Plasticity for Anomaly Detection
0 users, 1 mentions 2020/12/22 05:21
[2006.07743] 3DFCNN: Real-Time Action Recognition using 3D Deep Neural Networks with Raw Depth Infor...
Read more [2006.07743] 3DFCNN: Real-Time Action Recognition using 3D Deep Neural Networks with Raw Depth Infor...
0 users, 1 mentions 2021/04/25 13:50
Divide-and-Assemble: Learning Block-Wise Memory for Unsupervised Anomaly Detection
Read more Divide-and-Assemble: Learning Block-Wise Memory for Unsupervised Anomaly Detection
0 users, 1 mentions 2021/10/24 22:37
GitHub - pfnet-research/FSCS: Fast Soft Color Segmentation
Read more GitHub - pfnet-research/FSCS: Fast Soft Color Segmentation
1 users, 1 mentions 2022/01/04 06:08
GitHub - soumith/ganhacks: starter from "How to Train a GAN?" at NIPS2016
Read more GitHub - soumith/ganhacks: starter from "How to Train a GAN?" at NIPS2016
22 users, 1 mentions 2022/01/05 15:09
Towards Driving-Oriented Metric for Lane Detection Models
Read more Towards Driving-Oriented Metric for Lane Detection Models
0 users, 1 mentions 2022/06/27 12:09

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Kaggleのカテゴリを新設しました