Enhancing Object-Attribute Alignment in Diffusion Models via Training-Free Contrastive Parallel Denoising
Bibliographic record
Abstract
Diffusion models (DMs) excel at image generation but often struggle with semantic misalignment in complex prompts involving multiple objects and attributes. Such challenges, including object missing, object-object fusion, and object-attribute misalignment, arise from inter-contamination in text embeddings during the denoising process. To address these issues, we propose Contrastive Parallel Denoising (CPD), a training-free approach that segments prompt into sub-prompts, encodes them independently, applies parallel denoising supported by novel intra-/inter- contrastive object-attribute losses to enhance spatial separation, and finally merges denoised latents using cross-attention maps. Evaluated on CompBench, CPD significantly outperforms state-of-the-art training-free diffusion methods, achieving superior results in both BLIP-VQA score and human evaluations. This work provides a robust framework for mitigating semantic misalignment in text-to-image generation. The code has been published at github.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.004 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".