Supplementary for Project RISE: Recognizing Industrial Smoke Emissions

The supplementary materials contain model weights for the RGB-I3D and RGB-TC models for the following paper:Yen-Chia Hsu, Ting-Hao (Kenneth) Huang, Ting-Yao Hu, Paul Dille, Sean Prendi, Ryan Hoffman, Anastasia Tsuhlares, Jessica Pachuta, Randy Sargent, and Illah Nourbakhsh. 2021. Project RISE: Recognizing Industrial Smoke Emissions. Proceedings of the AAAI Conference on Artificial Intelligence (AAAI 2021).Check the following GitHub repository about how to use the model weights:https://github.com/CMU-CREATE-Lab/deep-smoke-machineTwo folders in this repository, RGB-I3D and RGB-TC, correspond to two models in the paper. Inside each model, there are splits from S0 to S5, corresponding to the train/validation/test splits in the paper. Inside each split, there are three folders: metadata, model, and viz. The metadata folder shows the videos that are used for each split. The model folder contains the best model weights in the experiment. The viz folder contains visualizations of true positive, false positive, true negative, and false negative examples.本补充材料包含对应下述论文的RGB-I3D与RGB-TC模型权重：作者为Yen-Chia Hsu、Ting-Hao (Kenneth) Huang、Ting-Yao Hu、Paul Dille、Sean Prendi、Ryan Hoffman、Anastasia Tsuhlares、Jessica Pachuta、Randy Sargent及Illah Nourbakhsh，论文题为《Project RISE：工业烟雾排放识别》，发表于2021年的《AAAI人工智能大会论文集》（AAAI 2021）。请查阅下述GitHub仓库以了解模型权重的使用方法：https://github.com/CMU-CREATE-Lab/deep-smoke-machine。该仓库内的RGB-I3D与RGB-TC两个文件夹分别对应论文中的两款模型。每个模型下包含从S0至S5的数据拆分（split），对应论文中的训练/验证/测试拆分。每个数据拆分下均设有metadata、model与viz三个子文件夹：metadata文件夹用于展示各数据拆分所使用的视频；model文件夹包含实验中表现最优的模型权重；viz文件夹则收录真阳性、假阳性、真阴性以及假阴性样本的可视化结果。