Pith. sign in

REVIEW 10 cited by

Pretraining is All You Need for Image-to-Image Translation

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2205.12952 v1 pith:UPUTMVHG submitted 2022-05-25 cs.CV

Tengfei Wang , Ting Zhang , Bo Zhang , Hao Ouyang , Dong Chen , Qifeng Chen , Fang Wen This is my paper
classification cs.CV
keywords translationimage-to-imagetrainingdiffusiongenerationmodelneedpretraining
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We propose to use pretraining to boost general image-to-image translation. Prior image-to-image translation methods usually need dedicated architectural design and train individual translation models from scratch, struggling for high-quality generation of complex scenes, especially when paired training data are not abundant. In this paper, we regard each image-to-image translation problem as a downstream task and introduce a simple and generic framework that adapts a pretrained diffusion model to accommodate various kinds of image-to-image translation. We also propose adversarial training to enhance the texture synthesis in the diffusion model training, in conjunction with normalized guidance sampling to improve the generation quality. We present extensive empirical comparison across various tasks on challenging benchmarks such as ADE20K, COCO-Stuff, and DIODE, showing the proposed pretraining-based image-to-image translation (PITI) is capable of synthesizing images of unprecedented realism and faithfulness.

Discussion (0). Sign in to comment.

Forward citations

Cited by 10 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Domain Transfer Becomes Identifiable via a Single Alignment

    cs.LG 2026-05 unverdicted novelty 7.0 of 10

    Domain transfer becomes identifiable from marginals plus one anchor under Jacobian sparsity, enabled by a randomized masked finite-difference regularizer.

  2. InstancePin: Instance-Addressable Layout-to-Image Diffusion via Coordinate Pinning

    cs.CV 2026-08 conditional novelty 6.0 of 10

    InstancePin adds per-instance coordinate tokens and mask-guided fusion to a frozen layout-to-image diffusion model, improving FID and mIoU on Cityscapes while reducing visual blending of nearby same-category objects.

  3. LooseControlVideo: Directorial Video Control using Spatial Blocking

    cs.CV 2026-06 unverdicted novelty 6.0 of 10

    LooseControlVideo fine-tunes a video model on DNOCS-annotated data to enable layout and trajectory control via oriented 3D boxes, reporting 1.2-3x gains in trajectory accuracy over 2D baselines on nuScenes, HO-3D and BEHAVE.

  4. Colorful-Noise: Training-Free Low-Frequency Noise Manipulation for Color-Based Conditional Image Generation

    cs.CV 2026-05 unverdicted novelty 6.0 of 10

    A training-free technique manipulates low-frequency noise in diffusion models to control image color and structure using low-frequency priors.

  5. Colorful-Noise: Training-Free Low-Frequency Noise Manipulation for Color-Based Conditional Image Generation

    cs.CV 2026-05 unverdicted novelty 6.0 of 10

    Low-frequency noise manipulation with image priors enables training-free conditioning of color and global structure in text-to-image diffusion models while preserving detail variability.

  6. HG-Lane: High-Fidelity Generation of Lane Scenes under Adverse Weather and Lighting Conditions without Re-annotation

    cs.CV 2026-03 conditional novelty 6.0 of 10

    HG-Lane synthesizes 30,000 adverse-weather lane images without re-annotation and boosts CLRNet mF1 by 20.87% on the resulting benchmark across normal, snow, rain, fog, night, and dusk conditions.

  7. T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models

    cs.CV 2023-02 unverdicted novelty 6.0 of 10

    T2I-Adapters are lightweight modules that enable fine-grained control over color and structure in text-to-image diffusion models by aligning external conditions with the frozen model's internal knowledge.

  8. eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers

    cs.CV 2022-11 unverdicted novelty 6.0 of 10

    An ensemble of stage-specialized text-to-image diffusion models improves prompt alignment over single shared-parameter models while preserving visual quality and inference speed.

  9. Generation of Heterogeneous PET Images from Uniform Organ Activity Maps Using a Pretrained Domain-Adapted Diffusion Model

    cs.CV 2026-05 unverdicted novelty 5.0 of 10

    A domain-adapted diffusion model synthesizes heterogeneous PET images from uniform organ activity maps, achieving high quantitative accuracy (CCC > 0.92) and visual realism comparable to real scans.

  10. Translationese as a Rational Response to Translation Task Difficulty

    cs.CL 2026-03 unverdicted novelty 5.0 of 10

    Translationese is partly predictable from quantifiable translation-task difficulty, especially cross-lingual transfer load, more so for English-to-German than the reverse.

Pith tools