AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

From Show to Tell: A Survey on Deep Learning-Based Image Captioning

Matteo Stefanini, Marcella Cornia, Lorenzo Baraldi, Silvia Cascianelli, Giuseppe Fiameni, Rita Cucchiara

TPAMI 2022 · Vision-language and multimodal › Human and embodied modeling · approach: Concept-guided image captioning · 450 citations (OpenAlex)

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 changedWhereMeasuredEffect
Replace global CNN features with region-based visual encoding (Up-Down vs Show, Attend and Tell)LSTM captioner · COCO Karpathy testCIDEr: 104.6 -> 122.7Improved
Add reinforcement learning with CIDEr reward (SCST FC vs Show and Tell)FC captioner · COCO Karpathy testCIDEr: 97.2 -> 104.5Improved
Add inter-object relations via graphs (SGAE) over region-only Up-DownUp-Down · COCO Karpathy testCIDEr: 122.7 -> 129.1Improved
Add vision-and-language pre-training (VinVL vs X-Transformer)Transformer captioner · COCO Karpathy testCIDEr: 132.8 -> 140.9Improved
Add vision-and-language pre-training (VinVL vs X-Transformer)Transformer captioner · COCO Karpathy testSPICE: 23.4 -> 25.2Improved
Increase model size (VinVL 369.6M vs X-LAN 75.2M parameters)captioner · COCO Karpathy testCIDEr: 132.0 -> 140.9Mixed
Use larger model (CPTR 138.5M vs M2 Transformer 38.4M parameters)Transformer captioner · COCO Karpathy testCIDEr: 131.2 -> 129.4Hurt
Use text-only BERT-S metric to discriminate approachesevaluation · COCO Karpathy testBERT-S: 88.5 -> 94.4No clear effect
Add self-attention (X-LAN X-Linear attention vs AoANet)LSTM captioner · COCO Karpathy testCIDEr: 129.8 -> 132.0Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy