InstanceFormer: An Online Video Instance Segmentation Framework

Authors: Rajat Koner, Tanveer Hannan, Suprosanna Shit, Sahand Sharifzadeh, Matthias Schubert, Thomas Seidl, Volker Tresp

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate Instance Former on three challenging datasets OVIS (Qi et al. 2021), YTVIS19 (Yang et al. 2019), and YTVIS-21 (Xu et al. 2021). ... Here, we analyze the quantitative and qualitative findings of our experiments across multiple dataset.
Researcher Affiliation Academia 1 Ludwig Maximilian University of Munich 2 Technical University of Munich 3 MCML
Pseudocode No The paper describes its methodology in narrative text and with equations, but it does not include a clearly labeled pseudocode block or algorithm.
Open Source Code Yes Code is available at https: //github.com/rajatkoner08/Instance Former.
Open Datasets Yes We evaluate Instance Former on three challenging datasets OVIS (Qi et al. 2021), YTVIS19 (Yang et al. 2019), and YTVIS-21 (Xu et al. 2021). ... We first pre-train our network on COCO (Lin et al. 2014) for 12 epochs.
Dataset Splits Yes We evaluate our model on the official validation split.
Hardware Specification Yes We train our network using the Adam W optimizer on 4 NVIDIA RTX A6000 GPUs with a batch size of 4 and learning rate of 1e 4 for 16 epochs.
Software Dependencies No The paper mentions using specific models and optimizers (e.g., 'Deformable-DETR', 'ResNet-50', 'Adam W optimizer') but does not specify software library version numbers (e.g., 'PyTorch 1.9') which would be needed for full reproducibility.
Experiment Setup Yes We train our network using the Adam W optimizer on 4 NVIDIA RTX A6000 GPUs with a batch size of 4 and learning rate of 1e 4 for 16 epochs. ... In our experiments, we have set the size of memory queue d = 4 and number of memory token per frame to 10.