Adversarial Alignment for Source Free Object Detection
Authors: Qiaosong Chu, Shuyan Li, Guangyi Chen, Kai Li, Xiu Li
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on multiple cross-domain object detection datasets demonstrate that our proposed method consistently outperforms the compared SFOD methods. |
| Researcher Affiliation | Collaboration | 1Tsinghua Shenzhen International Graduate School, Shenzhen, China 2Tsinghua University, Beijing, China 3Carnegie Mellon University, Pittsburgh PA, USA 4Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, UAE 5NEC LABORATORIES AMERICA, INC |
| Pseudocode | No | The paper describes its methods and processes but does not include any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | Our implementation is available at https://github.com/Chu Qiaosong. |
| Open Datasets | Yes | We evaluated our method on five popular object detection datasets. The detailed information of these datasets is summarized in the following: (1)Cityscapes (Cordts et al. 2016) ... (2)Foggy-Cityscapes (Sakaridis et al. 2018) ... (3)KITTI (Geiger, Lenz, and Urtasun 2012) ... (4)Sim10k (Johnson-Roberson et al. 2017) ... (5)BDD100k (Yu et al. 2018). |
| Dataset Splits | Yes | (1)Cityscapes (Cordts et al. 2016) ... which contains 2,975 training images and 500 validation images. ... (5)BDD100k (Yu et al. 2018) ... 36,728 images were used for training and the other 5,258 images for validation. |
| Hardware Specification | No | The paper does not provide specific details about the hardware (e.g., CPU, GPU models, memory) used for running the experiments. |
| Software Dependencies | Yes | All experiments were implemented with Py Torch 1.7.1. |
| Experiment Setup | Yes | The detector was trained with Stochastic Gradient Descent (SGD) with a learning rate of 0.001. ... we only updated the teacher model every 2500 iterations using exponential moving average (EMA) weights of the student model. In the pseudo label generation process, we filtered out the bounding boxes whose classification scores were lower than 0.7 to control the quality of pseudo labels. ... we set λ = 0.1 for Sim10k Cityscapes in Equ. (7) and λ = 1 for other tasks. We set the threshold parameter σ = 0.7 as it is empirically found to result in the best performance. |