RADIANT: Radar-Image Association Network for 3D Object Detection

Authors: Yunfei Long, Abhinav Kumar, Daniel Morris, Xiaoming Liu, Marcos Castro, Punarjay Chakravarty

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results show significant improvement in mean average precision and translation error on the nu Scenes dataset over monocular counterparts.
Researcher Affiliation Collaboration 1 Michigan State University 2 Ford Motor Company {longyunf, kumarab6, dmorris, liuxm}@msu.edu, {mgerard8, pchakra5}@ford.com
Pseudocode No The paper does not contain any pseudocode or algorithm blocks.
Open Source Code Yes Our source code is available at https://github.com/longyunf/radiant.
Open Datasets Yes We apply the proposed method on the detection task of nu Scenes dataset (Caesar et al. 2020), a widely used dataset with both image and radar points collected in urban driving environment.
Dataset Splits Yes The nu Scenes detection dataset consists of 28,130 training samples, 6,019 validation samples and 6,008 test samples.
Hardware Specification No The paper does not provide specific details about the hardware used for experiments (e.g., GPU models, CPU types).
Software Dependencies No The paper does not specify version numbers for any software dependencies, such as programming languages or libraries.
Experiment Setup No The paper mentions using FCOS3D and ResNet-18, and describes architectural choices like freezing the image branch, but it does not provide specific hyperparameters (e.g., learning rate, batch size, epochs) or detailed training configurations in the main text. It refers to supplementary material for 'Details of the input vector' for the DWN, but not for general experiment setup.