Manifold-Aware Self-Training for Unsupervised Domain Adaptation on Regressing 6D Object Pose
Authors: Yichen Zhang, Jiehong Lin, Ke Chen, Zelin Xu, Yaowei Wang, Kui Jia
IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiment results on three public benchmarks of the challenging 6D pose estimation task can verify the effectiveness of our method, consistently achieving superior performance to the state-of-the-art for UDA on 6D pose estimation. |
| Researcher Affiliation | Academia | 1South China University of Technology 2Peng Cheng Laboratory {eezyc, lin.jiehong, eexuzelin}@mail.scut.edu.cn, {chenk, kuijia}@scut.edu.cn, wangyw@pcl.ac.cn |
| Pseudocode | No | The paper describes the method using textual descriptions and pipeline figures, but it does not include structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | Code is available at https://github.com/Gorilla-Lab-SCUT/MAST. |
| Open Datasets | Yes | The Line MOD dataset [Hinterstoisser et al., 2012] provides individual videos of 13 textureless objects... The Occluded Line MOD dataset [Brachmann et al., 2014] is a subset of the Line MOD... The Homebrewed DB dataset [Kaskman et al., 2019] provides newly captured test images... We employ the publicly available synthetic data provided by BOP challenge [Hodaˇn et al., 2020] as the source data... |
| Dataset Splits | No | For each object, we follow [Brachmann et al., 2014] to use randomly sampled 15% of the sequence as the real-world training data of the target domain, and the remaining images are set aside for testing. While training and testing splits are mentioned, an explicit validation split with specific percentages or methodology is not detailed. |
| Hardware Specification | No | The paper does not provide specific details about the hardware (e.g., GPU or CPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions using the 'Adam optimizer' and 'Mask R-CNN' but does not specify any software versions for dependencies like Python, PyTorch, TensorFlow, or other libraries. |
| Experiment Setup | Yes | For pose estimation, we set the numbers of anchors as NR = 60, Nvx = Nvy = 20, Nz = 40, and set the ranges of vx, vy and z as [dmin vx , dmax vx ] = [dmin vy , dmax vy ] = [ 200, 200], and [dmin z , dmax z ] = [0.0, 2.0], respectively. ... Their initial learning rates are 3 10 4 and 3 10 5, respectively. The training batch size is set as B = 32. We also include the same data augmentation as [Labb e et al., 2020] during training. |