Semantic Object Accuracy for Generative Text-to-Image Synthesis
Explore this paper's citation graph
Summary
A new model that explicitly models individual objects within an image and a new evaluation metric called Semantic Object Accuracy (SOA) that specifically evaluates images given an image caption are introduced that outperform models which only model global image characteristics.
- Type
- article
- Published
- 2019-10-29
- Cited by
- 188
- References
- 67
- Access
- Open access
- OpenAlex
- https://openalex.org/W2982450728
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:204949374
Keywords
Computer science, Artificial intelligence, Generative grammar, Natural language processing, Object (grammar)
References
- Linguistic Features for Automatic Evaluation of Heterogenous MT Systems
- Conditional Generative Adversarial Nets
- Generative Adversarial Text to Image Synthesis
- Improved Techniques for Training GANs
- Show and Tell: Lessons Learned from the 2015 MSCOCO Image Captioning Challenge
- Learning What and Where to Draw
- Learning to Generate Images of Outdoor Scenes from Attributes and Semantic Layouts
- StackGAN: Text to Photo-Realistic Image Synthesis with Stacked Generative Adversarial Networks
- Photographic Image Synthesis with Cascaded Refinement Networks
- Generating Interpretable Images with Controllable Structure
- StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks
- A Note on the Inception Score
- Inferring Semantic Layout for Hierarchical Text-to-Image Synthesis
- Pros and Cons of GAN Evaluation Measures
- ChatPainter: Improving Text to Image Generation using Dialogue
- YOLOv3: An Incremental Improvement
- Learning Hierarchical Semantic Image Manipulation through Structured Representations
- Probabilistic Neural Programmed Networks for Scene Generation
- Adversarial Learning of Semantic Relevance in Text to Image Synthesis
- LayoutGAN: Generating Graphic Layouts with Wireframe Discriminators
Cited by
- Instance Mask Embedding and Attribute-Adaptive Generative Adversarial Network for Text-to-Image Synthesis
- Learning Layout and Style Reconfigurable GANs for Controllable Image Synthesis
- TIME: Text and Image Mutual-Translation Adversarial Networks
- Video Content Understanding Using Text
- Improving Text to Image Generation using Mode-seeking Function
- Leveraging Visual Question Answering to Improve Text-to-Image Synthesis
- Using Text to Teach Image Retrieval
- Attributes Aware Face Generation with Generative Adversarial Networks
- TR at SemEval-2020 Task 4: Exploring the Limits of Language-model-based Common Sense Validation
- Adversarial Text-to-Image Synthesis: A Review
- CharacterGAN: Few-Shot Keypoint Character Animation and Reposing
- AttrLostGAN: Attribute Controlled Image Synthesis from Reconfigurable Layout and Style
- GUIGAN: Learning to Generate GUI Designs Using Generative Adversarial Networks
- Image generation from text with entity information fusion
- Cross-Modal Contrastive Learning for Text-to-Image Generation
- Video Generation from Text Employing Latent Path Construction for Temporal Modeling
- FA-GAN: Feature-Aware GAN for Text to Image Synthesis
- Transformer models for enhancing AttnGAN based text to image generation
- A Survey on Multimodal Deep Learning for Image Synthesis: Applications, methods, datasets, evaluation metrics, and results comparison
- Image Synthesis from Layout with Locality-Aware Mask Adaption
Related papers
- A survey of image synthesis and editing with generative adversarial networks
- ACGAN: Attribute controllable person image synthesis GAN for pose transfer
- Relationships between Image Synthesis and Analysis Towards Unification?
- Aesthetic-Aware Text to Image Synthesis
- GAN-based models and applications
- Progressive Semantic Image Synthesis via Generative Adversarial Network
- Language-Based Image Editing with Recurrent Attentive Models
- Decorating Your Own Bedroom: Locally Controlling Image Generation with Generative Adversarial Networks