Pith. sign in

REVIEW 16 cited by

Re-imagine the Negative Prompt Algorithm: Transform 2D Diffusion into 3D, alleviate Janus problem and Beyond

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 2304.04968 v3 pith:EOQUB3BU submitted 2023-04-11 cs.CV cs.GRcs.LG

Mohammadreza Armandpour , Ali Sadeghian , Huangjie Zheng , Amir Sadeghian , Mingyuan Zhou This is my paper
classification cs.CVcs.GRcs.LG
keywords perp-negimagesnegativealgorithmdiffusionmodelproblemprompts
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Although text-to-image diffusion models have made significant strides in generating images from text, they are sometimes more inclined to generate images like the data on which the model was trained rather than the provided text. This limitation has hindered their usage in both 2D and 3D applications. To address this problem, we explored the use of negative prompts but found that the current implementation fails to produce desired results, particularly when there is an overlap between the main and negative prompts. To overcome this issue, we propose Perp-Neg, a new algorithm that leverages the geometrical properties of the score space to address the shortcomings of the current negative prompts algorithm. Perp-Neg does not require any training or fine-tuning of the model. Moreover, we experimentally demonstrate that Perp-Neg provides greater flexibility in generating images by enabling users to edit out unwanted concepts from the initially generated images in 2D cases. Furthermore, to extend the application of Perp-Neg to 3D, we conducted a thorough exploration of how Perp-Neg can be used in 2D to condition the diffusion model to generate desired views, rather than being biased toward the canonical views. Finally, we applied our 2D intuition to integrate Perp-Neg with the state-of-the-art text-to-3D (DreamFusion) method, effectively addressing its Janus (multi-head) problem. Our project page is available at https://Perp-Neg.github.io/

Discussion (0). Sign in to comment.

Forward citations

Cited by 16 Pith papers

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

  1. Variance Reduction on the Camera Axis: Multi-View Score Distillation for 3D

    cs.CV 2026-06 unverdicted novelty 7.0 of 10

    MV-SDI aggregates K-view gradients per step via accumulation and antithetic pairs at fixed UNet budget, raising CLIP R-Precision from 74.8% to 83.8% (K=2) and halving steps while keeping the 2D prior frozen.

  2. How to Spin an Object: First, Get the Shape Right

    cs.CV 2024-12 unverdicted novelty 7.0 of 10

    Camera-Relative Object Coordinates (CROCS) as an intermediate geometry representation in two-stage image-to-3D models yields superior novel-view quality, geometric accuracy, and multiview consistency over depth maps, ...

  3. DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation

    cs.CV 2023-09 unverdicted novelty 7.0 of 10

    DreamGaussian creates high-quality textured 3D meshes from single-view images in 2 minutes via generative Gaussian Splatting with mesh extraction and UV refinement.

  4. Hallo4D: Multi-Modal Hallucination Mitigation for Consistent Spatio-Temporal Generation

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Hallo4D uses vision-language models to detect and correct spatial and temporal mistakes in AI-generated 3D and 4D content, improving consistency without retraining the base generators.

  5. Probability-Flow Distillation: Exact Wasserstein Gradient Flow for High-Fidelity 3D Generation

    cs.CV 2026-05 unverdicted novelty 6.0 of 10

    Probability-Flow Distillation exactly matches the Wasserstein gradient flow of the target distribution when distilling 2D diffusion priors into 3D models, yielding higher-fidelity results than SDS or SDI.

  6. CLOAK: Contrastive Guidance for Latent Diffusion-Based Data Obfuscation

    cs.LG 2025-12 conditional novelty 6.0 of 10

    CLOAK uses contrastive-learned public-attribute representations and negated classifier guidance to steer a latent diffusion model, producing obfuscated sensor data that preserves utility while suppressing private attributes.

  7. Enhancing Physical Plausibility in Video Generation by Reasoning the Implausibility

    cs.CV 2025-09 unverdicted novelty 6.0 of 10

    A training-free framework uses physics-violating counterfactual prompts and Synchronized Decoupled Guidance to suppress implausible motions in diffusion-based video generation while preserving photorealism.

  8. Dual Orthogonal Guidance for Robust Diffusion-based Handwritten Text Generation

    cs.CV 2025-08 unverdicted novelty 6.0 of 10

    A new guidance method for diffusion-based text generation subtracts the orthogonal component of a negative prompt from a positive prompt to reduce artifacts and increase style variation.

  9. SyncDreamer: Generating Multiview-consistent Images from a Single-view Image

    cs.CV 2023-09 unverdicted novelty 6.0 of 10

    SyncDreamer produces multiview-consistent images from a single input image by jointly modeling their distribution and synchronizing intermediate diffusion states via 3D-aware attention.

  10. Hallo4D: Multi-Modal Hallucination Mitigation for Consistent Spatio-Temporal Generation

    cs.CV 2026-07 conditional novelty 5.0 of 10

    Hallo4D mitigates 3D/4D generation hallucinations via LMM-based detection, multi-model voting correction, and motion-aware optimization without retraining base generators.

  11. Bootstrap Your Generator: Unpaired Visual Editing with Flow Matching

    cs.CV 2026-06 unverdicted novelty 5.0 of 10

    ByG enables unpaired training of flow matching editing models by pairing self-extracted instruction-following cues with cycle-consistency and routing gradients from clean predictions to noisy states.

  12. Structural Energy Guidance for View-Consistent Text-to-3D Generation

    cs.CV 2026-05 unverdicted novelty 5.0 of 10

    SEGS constructs structural energy in the PCA subspace of U-Net features and injects its gradient into the denoising process to improve multi-view consistency in text-to-3D generation.

  13. EmDT: Embedding Diffusion Transformer for Tabular Data Generation in Fraud Detection

    stat.ML 2026-03 unverdicted novelty 5.0 of 10

    EmDT combines UMAP clustering with a Transformer-based diffusion process to create synthetic fraud samples that improve XGBoost classification on credit card fraud data while preserving correlations and privacy.

  14. Few-step Flow for 3D Generation via Marginal-Data Transport Distillation

    cs.CV 2025-09 conditional novelty 5.0 of 10

    MDT-dist distills a pretrained 3D flow model into a 1-2 step generator using velocity matching plus velocity distillation, cutting TRELLIS inference from 6.1s to 0.68s while approximately preserving generation quality.

  15. ConsDreamer: Advancing Multi-View Consistency for Zero-Shot Text-to-3D Generation

    cs.CV 2025-04 unverdicted novelty 5.0 of 10

    ConsDreamer refines conditional and unconditional terms in score distillation via view disentanglement and geometric consistency loss to reduce the Janus problem in zero-shot text-to-3D.

  16. Advances in 4D Representation: Geometry, Motion, and Interaction

    cs.CV 2025-10 conditional novelty 4.0 of 10

    A representation-centric survey of 4D generation and reconstruction, organized by geometry, motion, and interaction, with qualitative trade-off comparisons across seven representation families.

Pith tools