Asymmetric Hierarchical Difference-aware Interaction Network for Event-guided Motion Deblurring

Authors: Wen Yang, Jinjian Wu, Leida Li, Weisheng Dong, Guangming Shi

AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on both synthetic and realworld datasets demonstrate that our method achieves state-of-the-art performance. Code https://github.com/wyang-vis/AHDINet Our AHDINet is evaluated on 1) Synthetic dataset. Go Pro (Nah, Hyun Kim, and Mu Lee 2017) and DVD (Su et al. 2017) datasets are widely adopted for imageonly and event-based deblurring such as (Sun et al. 2022), which contains synthetic blurring images and sharp clear ground-truth images, as well as synthetic events generated by simulation algorithm ESIM (Rebecq, Gehrig, and Scaramuzza 2018). 2) Authentic dataset. REBlur (Sun et al. 2022) is a genuine event deblurring dataset collected by DAVIS, with an image resolution of 360 260. To evaluate the effectiveness of the key components (EEC and ISC) in our model, we conduct ablation studies on Go Pro dataset.
Researcher Affiliation Academia 1School of Artificial Intelligence, Xidian University, Xi an 710071, China 2Pazhou Lab, Huangpu, 510555, China EMAIL, EMAIL
Pseudocode No The paper describes the proposed modules (EEC and ISC) in detail using mathematical formulations and descriptive text, but it does not contain any clearly labeled pseudocode or algorithm blocks.
Open Source Code Yes Code https://github.com/wyang-vis/AHDINet
Open Datasets Yes Our AHDINet is evaluated on 1) Synthetic dataset. Go Pro (Nah, Hyun Kim, and Mu Lee 2017) and DVD (Su et al. 2017) datasets are widely adopted for imageonly and event-based deblurring such as (Sun et al. 2022), which contains synthetic blurring images and sharp clear ground-truth images, as well as synthetic events generated by simulation algorithm ESIM (Rebecq, Gehrig, and Scaramuzza 2018). 2) Authentic dataset. REBlur (Sun et al. 2022) is a genuine event deblurring dataset collected by DAVIS, with an image resolution of 360 260.
Dataset Splits Yes Authentic dataset. REBlur (Sun et al. 2022) is a genuine event deblurring dataset collected by DAVIS, with an image resolution of 360 260. This dataset comprises 1,389 sample pairs encompassing diverse 12 distinct types of linear and nonlinear motions, for three different moving patterns and the camera itself. Among these samples, there are 486 training sets and 903 test sets.
Hardware Specification Yes Our method is implemented using Pytorch on NVIDIA RTX 3090 GPU.
Software Dependencies No Our method is implemented using Pytorch on NVIDIA RTX 3090 GPU. The paper mentions PyTorch but does not specify its version number, nor does it list any other software dependencies with specific versions.
Experiment Setup Yes The size of training patch is 300 300 with minibatch size of 8. The optimizer is ADAM (Kingma and Ba 2015), and the learning rate is initialized at 2 10 4 and decreased by the cosine learning rate strategy with a minimum learning rate of 10 6. For data augmentation, each patch is horizontally flipped with the probability of 0.5. The training ends after 200k iterations for Go Pro dataset and 100k iterations for REBlur dataset.