← Back to the feed
Energy & materialsResearch findings

A Novel CTR-Driven Advertising Image Generation Method Using Multimodal LLMs

Original published 2025.02.05 18:06 KSTAdded to Blackdeck 2026.10.24 19:51 KST

What has changed?

A multimodal large language model is proposed to generate advertising images optimized for click-through rate (CTR)

Key points

  1. Advertising images are generated to optimize user click-through rate (CTR) as the primary objective
  2. Multimodal large language models are pre-trained and fine-tuned using a reward model based on reinforcement learning
  3. A product-centric preference optimization strategy ensures generated backgrounds align with product characteristics

Connect it to your work

A new approach to enhance user engagement and ad effectiveness through CTR-based image generation

What remains unproven

This research note is based on an abstract; check the original for test conditions and application scope.

Originals used for this note

Links open the original publisher in a new tab.

  1. arxivCTR-Driven Advertising Image Generation with Multimodal Large Language Models