Generative Modeling of Infinite Occluded Objects for Compositional Scene Representation

Authors: Jinyang Yuan, Bin Li, Xiangyang Xue

ICML 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments conducted on a series of specially designed datasets demonstrate that the proposed method outperforms two state-of-the-art methods when object occlusions exist.
Researcher Affiliation Academia 1Shanghai Key Laboratory of Intelligent Information Processing; Fudan-Qiniu Joint Laboratory for Deep Learning; Shanghai Institute of Intelligent Electronics & Systems; School of Computer Science, Fudan University, China.
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide any information about open-source code for the described methodology.
Open Datasets Yes The perceptual grouping performance of the compared methods are evaluated on a series of datasets derived from the publicly released datasets provided by (Greff et al., 2016b;a; 2017). The size of images in all datasets is 48 48, and each image may contain 2 4 binary hollow shapes (referred as Shapes) or real-valued handwritten digits (referred as MNIST).
Dataset Splits Yes In all datasets, 50,000, 10,000, and 10,000 images are used for training, validation, and test, respectively.
Hardware Specification No The paper does not explicitly describe the hardware used for running its experiments.
Software Dependencies No The paper does not provide specific version numbers for software dependencies.
Experiment Setup Yes All three methods are trained on images containing 2 or 3 objects with K =4.