Research map / Image generation / GANs
Contextual RNN-GANs for Abstract Reasoning Diagram Generation
Viveka Kulharia, Arnab Ghosh, Amitabha Mukerjee, Vinay Namboodiri, Mohit Bansal
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the Context-RNN-GAN with Siamese CNN features instead of RNN-GAN with the same features | RNN-GAN · DAT-DAR | accuracy: 31.2% -> 35.4% | Improved |
| Use the Context-RNN-GAN with Siamese CNN features instead of a regular RNN with Siamese CNN features and L1 loss | Regular RNN · DAT-DAR | accuracy: 30.8 % -> 35.4% | Improved |
| Add contextual history to the discriminator (ablation) | RNN-GAN with Siamese CNN features · DAT-DAR | accuracy: 31.2% -> 35.4% | Improved |
| Use Siamese CNN features instead of raw pixels (ablation) | Context-RNN-GAN · DAT-DAR | accuracy: 30.3% -> 35.4% | Improved |
| Use Siamese CNN features instead of HOG features (ablation) | Context-RNN-GAN · DAT-DAR | accuracy: 34.1% -> 35.4% | Improved |
| Use Siamese CNN features instead of Shallow CNN features (ablation) | Context-RNN-GAN · DAT-DAR | accuracy: 33.0% -> 35.4% | Improved |
| Use Context-RNN-GAN instead of AE-Conv-LSTM without optical flow | AE-Conv-LSTM (w/o optical flow) · Moving-MNIST | cross-entropy error: 262.6 -> 241.8 | Improved |
| Use Context-RNN-GAN instead of LSTM | LSTM · Moving-MNIST | cross-entropy error: 341.2 -> 241.8 | Improved |
| Use Context-RNN-GAN instead of LSTM | LSTM · Moving-MNIST | squared error: 208.1 -> 167.9 | Improved |
Closest papers by meaning
- Rectifying GAN Gradient Supervision Via Score Functions (TPAMI 2026 · similarity 0.85)
- Generating Triples With Adversarial Networks for Scene Graph Construction (AAAI 2018 · similarity 0.85)
- Toward a Visual Concept Vocabulary for GAN Latent Space (ICCV 2021 · similarity 0.85)
- Dynamic Graph Generation Network: Generating Relational Knowledge From Diagrams (CVPR 2018 · similarity 0.85)
- Boundary Seeking GANs (ICLR 2018 · similarity 0.85)
- RNNs are not Transformers (Yet): The Key Bottleneck on In-Context Retrieval (ICLR 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
