MSDN
Scene graph generator
An implementation of a multi-level scene description network in PyTorch for generating scene graphs from object, phrase, and region captions.
This is our PyTorch implementation of Multi-level Scene Description Network (MSDN) proposed in our ICCV 2017 paper.
227 stars
13 watching
51 forks
Language: Python
last commit: about 5 years ago Related projects:
| Repository | Description | Stars |
|---|---|---|
 yuweihao/kern | An open-source implementation of a graph neural network architecture for scene graph generation in computer vision | 120 |
 jwyang/graph-rcnn.pytorch | Reimplemented scene graph generation models using PyTorch 1.0 | 733 |
 kenneth-wong/scenegraph_mem | Research software implementing scene graph generation methods from CVPR 2019 paper | 4 |
 kaihuatang/vctree-scene-graph-generation | Generates scene graphs from visual contexts using dynamic tree structures | 121 |
| kaihuatang/scene-graph-benchmark.pytorch | A PyTorch implementation of Scene Graph Generation methods with support for visualization and evaluation on custom images | 1,075 |
 ashual/scene_generation | A PyTorch implementation of a deep learning-based method for generating interactive scenes with specified object attributes and relations | 187 |
 yangxuntu/sgae | Automatically generates scene graphs from images to aid in image captioning tasks | 220 |
| kenneth-wong/het-eccv20 | Generates human-mimetic hierarchical scene graphs from images using a hierarchical approach. | 16 |
 siml3/gps-net | A system for generating scene graphs from video data using a graph property sensing network architecture | 63 |
 jayleicn/animegan | A PyTorch implementation of Generative Adversarial Networks for anime face drawing | 1,280 |
 vincentschen/limited-label-scene-graphs | This project develops a semi-supervised method to generate training datasets for scene graphs using limited labels and unlabeled data. | 54 |
 yiwuzhong/sub-gc | A PyTorch implementation of image captioning models via scene graph decomposition. | 96 |
 casualganpapers/make-a-scene | PyTorch implementation of a scene-based text-to-image generation system with human priors using Generative Adversarial Networks and semantic segmentation | 333 |
 hanzhanggit/stackgan | A PyTorch implementation of a generative adversarial network for image synthesis from text descriptions | 1,860 |
 mfessenden/scenegraph | A UI framework for visualizing node graphs in visual effects applications. | 162 |