Deep Partial Rank Aggregation for Personalized Attributes
Authors: Qianqian Xu, Zhiyong Yang, Zuyao Chen, Yangbangyan Jiang, Xiaochun Cao, Yuan Yao, Qingming Huang678-688
AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In our empirical studies, we perform a series of experiments on three real-world datasets: LFW-10, Shoes, and Sun. The corresponding results consistently show the superiority of our proposed model. |
| Researcher Affiliation | Academia | Qianqian Xu 1, Zhiyong Yang 2,3, Zuyao Chen 4, Yangbangyan Jiang 2,3, Xiaochun Cao 2,3,6, Yuan Yao 7, Qingming Huang 1,4,5,6 1 Key Laboratory of Intelligent Information Processing, Institute of Computing Technology, CAS, Beijing, China 2 State Key Laboratory of Information Security, Institute of Information Engineering, CAS, Beijing, China 3 School of Cyber Security, University of Chinese Academy of Sciences, Beijing, China 4 School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing, China 5 Key Laboratory of Big Data Mining and Knowledge Management, Chinese Academy of Sciences, Beijing, China 6 Peng Cheng Laboratory, Shenzhen, China 7 Department of Mathematics, Hong Kong University of Science and Technology, Hong Kong, China |
| Pseudocode | No | The paper does not contain any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain any explicit statements about making the source code for their methodology publicly available or provide a link to a code repository. |
| Open Datasets | Yes | In our empirical studies, we perform a series of experiments on three real-world datasets: LFW-10, Shoes, and Sun. The LFW-10 dataset (Sandeep, Verma, and Jawahar 2014) consists of 2,000 face images, which are chosen from the Labeled Faces in the Wild (Huang et al. 2008) dataset. The Shoes dataset is collected from (Kovashka and Grauman 2015). The SUN Attribute dataset is a well-known large-scale scene attribute dataset. |
| Dataset Splits | Yes | LFW-10: The images are split to 1000/1000 to create training/testing pairs. The resulting dataset has 50,000 annotated sample pairs, with 500 training and testing pairs per attribute. |
| Hardware Specification | No | The paper does not provide specific details about the hardware (e.g., GPU models, CPU types, memory) used to run the experiments. |
| Software Dependencies | No | The paper mentions: "We implement the models using library Pytorch (Paszke et al. 2019)". While Pytorch is named, a specific version number (e.g., PyTorch 1.5) is not provided, making it not reproducible based on this information alone. |
| Experiment Setup | Yes | For training, we use a mini-batch size of 128 image pairs for SGD. We set the initial learning rate to 1e-3 and fix the momentum to 0.9. We train these networks for 300 epochs, and the learning rate is reduced by a factor of 10 every 40 epochs. We use random crops of size 227x227 from our 256x256 input image during training and resize all images to 227x227 for testing. |