AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

PUG: Photorealistic and Semantically Controllable Synthetic Data for Representation Learning

Florian Bordes, Shashank Shekhar, Mark Ibrahim, Diane Bouchacourt, Pascal Vincent, Ari S. Morcos

NeurIPS 2023 · Image generation › Autoregressive and latent generation · approach: Synthetic training data from generators · 4 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
fine-tune CLIP on PUG:AR4T (200K)CLIP-ViT-B/32 · ARO averageaccuracy: 57.32 -> 65.36Improved
fine-tune CLIP on PUG:AR4T (1M instead of 200K)CLIP-ViT-B/32 · ARO averageaccuracy: 65.36 -> 67.52Improved
fine-tune CLIP on PUG:AR4T (200K)CLIP-ViT-B/32 · PUG: SPAR single animal left/rightPrecision@1: 49.84 -> 50.78No clear effect
fine-tune CLIP on PUG:AR4T (200K)CLIP-ViT-B/32 · PUG: SPAR double animal left/rightPrecision@1: 54.42 -> 54.23No clear effect
fine-tune CLIP on PUG:AR4T (1M instead of 200K)CLIP-ViT-B/32 · PUG: SPARPrecision@1: 50.78 -> 50.16No clear effect
fine-tune CLIP on PUG:AR4T instead of Syn-CLIPCLIP-ViT-B/32 · ARO averageaccuracy: 67.09 -> 65.36No clear effect
train on 30 backgrounds instead of 50 (ablation)ResNet50 · PUG: Animals held-out backgroundsaccuracy: 80 -> drops significantlyHurt
show every background for some animals (C>0) (ablation)ResNet50 · PUG: Animals held-out backgroundsaccuracyHurt
at least 10 animals seen with every texture (ablation)ResNet50 · PUG: Animals held-out texturesaccuracyImproved
train on medium size only vs small only (ablation)ResNet50 · PUG: Animals held-out sizesaccuracyImproved
higher ImageNet val accuracy (ViT-B32 21k vs Swin-B)ViTBasePretrained21k · PUG: ImageNettop-1 accuracy: 84.3 -> 83.6No clear effect
hard-negative caption instead of full caption setVLMs · PUG: SPARaccuracyHurt

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