AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Make a Cheap Scaling: A Self-Cascade Diffusion Model for Higher-Resolution Adaptation

Lanqing Guo, Yingqing HE, Haoxin Chen, Menghan Xia, Xiaodong Cun, Yufei Wang, Siyu Huang, Yong Zhang, Xintao Wang, Qifeng Chen, Ying Shan, Bihan Wen

ECCV 2024 · Image generation › Text-to-image diffusion · approach: Text-driven synthesis and semantic guidance · 37 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
add pivot-guided self-cascade (tuning-free) to SD 2.1 at 1024^2SD 2.1 · Laion-5B 1024^2FID_r: 29.89 -> 12.25Improved
add pivot-guided self-cascade (tuning-free) to SD 2.1 at 2048^2SD 2.1 · Laion-5B 2048^2FID_r: 104.70 -> 38.99Improved
tune lightweight upsampler modules (0.002M params) instead of full fine-tuningSD 2.1 · Laion-5B 2048^2FID_r: 43.55 -> 18.46Improved
tune upsampler modules instead of LoRA rank 4SD 2.1 · Laion-5B 1024^2FID_r: 14.74 -> 12.40Improved
add tuning of upsampler modules to the tuning-free cascade (ablation)Ours-TF · Laion-5B 1024^2FID_r: 12.25 -> 12.40No clear effect
add tuning of upsampler modules to the tuning-free cascade (ablation)Ours-TF · Laion-5B 1024^2pFID_r: 19.59 -> 15.35Improved
tuning-free cascade on extreme examples with fine-grained repeated objectsOurs-TF · qualitative failuresrepetition artifactsHurt
reduce training from 50k to 10k steps for full fine-tuning on video (ablation)Full Fine-tuning · Webvid-10MFVD_r: 531.57 -> 721.32Hurt
reduce LoRA training from 50k to 10k steps on video (ablation)LORA-R4 · Webvid-10MFVD_r: 623.72 -> 1221.46Hurt
use self-cascade tuning instead of SD+SR post-processingSD 2.1 · Laion-5B 1024^2pFID_r: 17.21 -> 15.35Improved

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.

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

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