reproducibilityindex.ai

Learning 1-Bit Tiny Object Detector with Discriminative Feature Refinement

Authors: Sheng Xu, Mingze Wang, Yanjing Li, Mingbao Lin, Baochang Zhang, David Doermann, Xiao Sun

Reproducibility Variable	Result	LLM Response
Research Type	Experimental	Extensive experiments on various tiny object detection (TOD) tasks demonstrate DFR-Det s superiority over state-of-the-art 1-bit detectors.
Researcher Affiliation	Collaboration	1Beihang University 2Skywork AI 3Zhongguancun Laboratory 4University at Buffalo 5Shanghai Artificial Intelligence Laboratory.
Pseudocode	No	The paper does not contain any pseudocode or algorithm blocks.
Open Source Code	No	The paper does not explicitly state that the source code for their methodology is available or provide a link.
Open Datasets	Yes	We evaluate the proposed method on AITOD (Wang et al., 2021), DOTA-v2.0 (Xia et al., 2018) and Tiny Person (Yu et al., 2020).
Dataset Splits	Yes	Models are trained on the AI-TOD trainval and validated on the AI-TOD test. Models are trained on the DOTA-v2.0 train and validated on the DOTA-v2.0 val.
Hardware Specification	Yes	We conduct the experiments on a computer with 1 NVIDIA RTX 3090 GPU, utilizing PyTorch (Paszke et al., 2019) for code construction.
Software Dependencies	No	The paper mentions 'PyTorch' but does not specify a version number or other software dependencies with version numbers.
Experiment Setup	Yes	All models are trained using the Stochastic Gradient Descent (SGD) optimizer for 12 epochs with 0.9 momentum, 0.0001 weight decay, and batch size of two. The initial learning rate is set to 0.005 and decays at the 8-th and 11-th epochs.