Norm-guided Adaptive Visual Embedding for Zero-Shot Sketch-Based Image Retrieval
Authors: Wenjie Wang, Yufeng Shi, Shiming Chen, Qinmu Peng, Feng Zheng, Xinge You
IJCAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on two challenging datasets demonstrate the superiority of our NAVE over state-of-the-art competitors. Extensive experimental results on two popular benchmarks demonstrate that our NAVE outperforms the state-of-the-arts by a significant margin. Table 1 presents m AP@all of sketch-based image retrieval (ZS-SBIR), sketch-based sketch retrieval (ZS-SBSR) and photo-based photo retrieval (ZS-PBPR) on unseen classes samples. In Table 4, we conduct ablation study to demonstrate the effectiveness of each component in our NAVE. |
| Researcher Affiliation | Academia | 1School of Electronic Information and Communication, Huazhong University of Science and Technology 2Shenzhen Research Institute of Huazhong University of Science and Technology 3Department of Computer Science and Engineering, Southern University of Science and Technology {wangwj5,yufengshi17, youxg}@hust.edu.cn |
| Pseudocode | Yes | Algorithm 1 NAVE algorithm |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for the described methodology is publicly available. |
| Open Datasets | Yes | We validate our NAVE on two widely-used benchmarks: Sketchy Ext. [Liu et al., 2017] and TU-Berlin Ext. [Zhang et al., 2016]. Following the data partitioning in [Liu et al., 2019], we randomly pick 25 classes from Sketchy and 30 classes from TU-Berlin as test set, and regard the rest 100/220 classes as the training set. |
| Dataset Splits | No | The paper specifies 'training set' and 'test set' but does not explicitly mention or detail a separate validation dataset split. For example: 'Following the data partitioning in [Liu et al., 2019], we randomly pick 25 classes from Sketchy and 30 classes from TU-Berlin as test set, and regard the rest 100/220 classes as the training set.' |
| Hardware Specification | Yes | We implement our NAVE with Py Torch and train it on two Nvidia 1080 Ti GPUs. |
| Software Dependencies | No | The paper mentions 'We implement our NAVE with Py Torch' (Section 5.1) but does not specify a version number for PyTorch or any other software dependencies with their respective versions. |
| Experiment Setup | Yes | We use Adam optimizer for training with an initial learning rate lr = 0.0001, β1 = 0.9, β2 = 0.999 and a 0.5 learning rate decay per epoch. The best hyper-parameters are λreg = 0.2, P = 5, τ = 3 for Sketchy Ext., and λreg = 0.1, P = 3, τ = 3 for TU-Berlin Ext. We follow the previous work [Liu et al., 2019] to take SE-Res Net50 pre-trained on Image Net as the feature extractor backbone and the same teacher network for a fair comparison. |