Semi-transductive Learning for Generalized Zero-Shot Sketch-Based Image Retrieval
Authors: Ce Ge, Jingyu Wang, Qi Qi, Haifeng Sun, Tong Xu, Jianxin Liao
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments are conducted on two large-scale benchmarks and four evaluation metrics. The results show that our method is superior over the state-of-the-art competitors in the challenging GZS-SBIR task. |
| Researcher Affiliation | Collaboration | State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing 100876, China {nwlgc, wangjingyu, qiqi8266, hfsun}@bupt.edu.cn, xutong@ebupt.com, jxlbupt@gmail.com |
| Pseudocode | No | The paper does not contain any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain an explicit statement about releasing source code for the described methodology, nor does it provide a link to a code repository. |
| Open Datasets | Yes | We employ two widely used SBIR datasets: Sketchy (Sangkloy et al. 2016) and TU-Berlin (Eitz, Hays, and Alexa 2012). |
| Dataset Splits | No | The paper mentions splitting datasets into seen and unseen classes and describes the generalized test set, but it does not specify the training/validation splits or percentages for the training phase, nor explicit details about a validation set beyond implying its use for early stopping. |
| Hardware Specification | No | The paper does not provide specific details about the hardware used for running its experiments, such as CPU/GPU models or memory specifications. |
| Software Dependencies | Yes | The whole model is implemented on top of Py Torch (Paszke et al. 2019) |
| Experiment Setup | Yes | The feature dimension of the embedding space is set to 1024D. The weighting factors for each dataset are determined by grid search with ω1 [0.01, 1] and ω2 [0.001, 10]. For Sketchy, ω1 = 0.5, ω2 = 0.1, and for TU-Berlin ω1 = 0.5, ω2 = 0.5. The margin hyperparameters in Lrank (Eq. 3) and Ltrans (Eq. 11) are empirically set to = 0.1 and δ = 0.01, respectively. The whole model is implemented on top of Py Torch (Paszke et al. 2019) and is trained end-to-end by stochastic gradient descent with learning rate 1e-3 and a mini-batch size 20. The early stopping strategy is adopted to combat overfitting. |