SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2308.09314 · ICCV · 2023

Retro-FPN: Retrospective Feature Pyramid Network for Point Cloud Semantic Segmentation

Xin Wen, Yu-Shen Liu, Yi Fang, Zhizhong Han, Hui Zhang, Peng Xiang

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Figure 1. (a) Encoder-decoder architecture with an inherent feature pyramid in the decoding stage. Black points with thicker outlines denote region features of larger local regions, green and purple points are the predicted semantic labels. (b) Retrospective Feature Pyramid Network, the points with orange outlines denote point-level semantic features. The rectangular areas highlighted in black and red denote local region feature learning and point-level semantic feature learning, respectively. In Retro-FPN, region information flows into points at all levels, and are retrospectively refined to the lowest level. (c) mIoU on S3DIS Area 5 with and without Retro-FPN.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2308.09314")
get_code_for_paper("2308.09314")
have("2308.09314")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2308.09314.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2308.09314.svg)](https://syntology.ai/paper/2308.09314)

Connect an agent — have() is free.