Research map / Image generation / Text-to-image diffusion
Turbo Learning for CaptionBot and DrawingBot
Qiuyuan Huang, Pengchuan Zhang, Dapeng Wu, Lei Zhang
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 |
|---|---|---|---|
| Jointly train CaptionBot and DrawingBot with turbo learning rather than training them separately | LSTM-XE CaptionBot and AttnGAN DrawingBot · COCO supervised, BLEU-4 reward | BLEU-4: 0.2684 -> 0.3168 | Improved |
| Jointly train CaptionBot and DrawingBot with turbo learning rather than training them separately | AttnGAN DrawingBot · COCO supervised, BLEU-4 reward | Inception score: 25.68 -> 26.69 | Improved |
| Jointly train CaptionBot and DrawingBot with turbo learning rather than training them separately | LSTM-XE CaptionBot · COCO supervised, CIDEr-D reward | CIDEr-D: 0.7362 -> 0.7473 | Improved |
| Jointly train CaptionBot and DrawingBot with turbo learning rather than training them separately | AttnGAN DrawingBot · COCO supervised, CIDEr-D reward | Inception score: 25.68 -> 26.72 | Improved |
| Use turbo learning with unlabeled samples and cycle-generated pseudo-labels | LSTM-XE CaptionBot · COCO semi-supervised, 1,000 labeled and 112,287 unlabeled training samples | BLEU-4: 0.1036 -> 0.1832 | Improved |
| Use turbo learning with unlabeled samples and cycle-generated pseudo-labels | LSTM-RL CaptionBot · COCO semi-supervised, 1,000 labeled and 112,287 unlabeled training samples | CIDEr-D: 0.4924 -> 0.6954 | Improved |
| Use the perceptual-like semantic reconstruction loss instead of pixel-wise or DSSM reconstruction loss (ablation) | Turbo learning for CaptionBot and DrawingBot · COCO supervised | performance | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AttnGAN — AttnGAN: Fine-Grained Text to Image Generation With Attentional Generative Adversarial Networks (CVPR 2018) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Teaching Machines to Describe Images with Natural Language Feedback (NeurIPS 2017 · similarity 0.84)
- From Show to Tell: A Survey on Deep Learning-Based Image Captioning (TPAMI 2022 · similarity 0.83)
- ChatGen: Automatic Text-to-Image Generation From FreeStyle Chatting (CVPR 2025 · similarity 0.83)
- Improved Image Captioning via Policy Gradient Optimization of SPIDEr (ICCV 2017 · similarity 0.83)
- Joint Optimization for Cooperative Image Captioning (ICCV 2019 · similarity 0.83)
- BalCapRL : A Balanced Framework for RL-Based MLLM Image Captioning (NeurIPS 2026 · similarity 0.82)
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.
