Joint COCO and Mapillary Workshop at ICCV 2019: COCO Instance Segmentation Challenge Track

10/06/2020
by   Zeming Li, et al.
0

In this report, we present our object detection/instance segmentation system, MegDetV2, which works in a two-pass fashion, first to detect instances then to obtain segmentation. Our baseline detector is mainly built on a new designed RPN, called RPN++. On the COCO-2019 detection/instance-segmentation test-dev dataset, our system achieves 61.0/53.1 mAP, which surpassed our 2018 winning results by 5.0/4.2 respectively. We achieve the best results in COCO Challenge 2019 and 2020.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
04/20/2023

Ensembling Instance and Semantic Segmentation for Panoptic Segmentation

We demonstrate our solution for the 2019 COCO panoptic segmentation task...
research
02/17/2022

Mirror-Yolo: An attention-based instance segmentation and detection model for mirrors

Mirrors can degrade the performance of computer vision models, however t...
research
04/18/2023

Perceive, Excavate and Purify: A Novel Object Mining Framework for Instance Segmentation

Recently, instance segmentation has made great progress with the rapid d...
research
02/02/2022

Automotive Parts Assessment: Applying Real-time Instance-Segmentation Models to Identify Vehicle Parts

The problem of automated car damage assessment presents a major challeng...
research
10/21/2020

2nd Place Solution to Instance Segmentation of IJCAI 3D AI Challenge 2020

Compared with MS-COCO, the dataset for the competition has a larger prop...
research
03/05/2018

Path Aggregation Network for Instance Segmentation

The way that information propagates in neural networks is of great impor...
research
12/17/2021

A Simple Single-Scale Vision Transformer for Object Localization and Instance Segmentation

This work presents a simple vision transformer design as a strong baseli...

Please sign up or login with your details

Forgot password? Click here to reset