MoNet3D: Towards Accurate Monocular 3D Object Localization in Real Time
Authors: Xichuan Zhou, Yicong Peng, Chunqiao Long, Fengbo Ren, Cong Shi
ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on the KITTI dataset show that the accuracy for predicting the depth and horizontal coordinates of objects in 3D space can reach 96.25% and 94.74%, respectively. Moreover, the method can realize the real-time image processing at 27.85 FPS |
| Researcher Affiliation | Academia | 1 Key Laboratory of Dependable Service Computing in Cyber Physical Society Ministry of Education, College of Microelectronics and Communication Engineering, Chongqing University, Chongqing, China 400044. 2 Arizona State University, Tempe, Arizona, United States. Correspondence to: Xichuan Zhou <zxc@cqu.edu.cn>. |
| Pseudocode | No | The paper describes the network structure and loss functions, but it does not include any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | Yes | Our code is publicly available at https://github. com/CQUlearningsystemgroup/ Yicong Peng. |
| Open Datasets | Yes | We performed experiments on the KITTI dataset to verify and evaluate the effectiveness of our algorithm. ... We used the same method as Chen to split KITTI data sets into 3712 training images and 3769 testing images(Chen et al., 2015). |
| Dataset Splits | No | The paper states 'We used the same method as Chen to split KITTI data sets into 3712 training images and 3769 testing images(Chen et al., 2015)', providing details for training and testing images. However, it does not explicitly specify the proportions or counts for a validation dataset split in the text, although 'KITTI validation dataset' is used in table headings. |
| Hardware Specification | Yes | Numerical experiments were performed on a computer equipped with an Inter Corei7-6900K CPU, 32GB of memory, and an NVIDIA Ge Force GTX 1080 Ti graphics card. |
| Software Dependencies | No | The paper mentions 'Model training uses tensorflow s SGD algorithm with momentum', but it does not specify a version number for TensorFlow or any other software dependencies. |
| Experiment Setup | Yes | The experimental hyperparameter settings referred to Mon GRNet. We initialized the model with random parameters. In the experiment, the similarity hyperparameter of Equation 1 was set to 100.00, and α, β and γ were all set to 10.00. Model training uses tensorflow s SGD algorithm with momentum, batchsize is set to 2 and learning rate is set to 10 5. A total of 800000 iterations were trained on the KITTI dataset. |