REVIEW 5 cited by
Divide & Bind Your Attention for Improved Generative Semantic Nursing
Not yet reviewed by Pith; the record is open.
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
Divide & Bind Your Attention for Improved Generative Semantic Nursing
read the original abstract
Emerging large-scale text-to-image generative models, e.g., Stable Diffusion (SD), have exhibited overwhelming results with high fidelity. Despite the magnificent progress, current state-of-the-art models still struggle to generate images fully adhering to the input prompt. Prior work, Attend & Excite, has introduced the concept of Generative Semantic Nursing (GSN), aiming to optimize cross-attention during inference time to better incorporate the semantics. It demonstrates promising results in generating simple prompts, e.g., "a cat and a dog". However, its efficacy declines when dealing with more complex prompts, and it does not explicitly address the problem of improper attribute binding. To address the challenges posed by complex prompts or scenarios involving multiple entities and to achieve improved attribute binding, we propose Divide & Bind. We introduce two novel loss objectives for GSN: a novel attendance loss and a binding loss. Our approach stands out in its ability to faithfully synthesize desired objects with improved attribute alignment from complex prompts and exhibits superior performance across multiple evaluation benchmarks.
Forward citations
Cited by 5 Pith papers
-
ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
ELLA introduces a timestep-aware semantic connector to link LLMs with diffusion models for improved dense prompt following, validated on a new 1K-prompt benchmark.
-
AI-T2I: Aggregating-and-Isolating Cross-Attention to Diffusion Models for Text-to-Image Synthesis
AI-T2I improves text-to-image alignment in diffusion models by using aggregation and isolation losses on cross-attention maps to fix scattering and overlap issues.
-
BindEdit: Taming Attention Leakage for Precise Multi-Object Image Editing
BindEdit suppresses two forms of attention leakage in diffusion-based editing by binding target tokens to regions, rebalancing cross-attention, and adding a region fidelity term, plus a new multi-object benchmark.
-
STEDiff: Strengthening Text Embedding for Text-to-Image Alignment in Diffusion Model
STEDiff improves semantic alignment in text-to-image diffusion models via training-free embedding strengthening with the [EOT] token and a spatial semantic loss, showing gains on T2I-CompBench.
-
DebFilter: Eradicating Biases Stashed in Value
DebFilter mitigates biases in text-to-image diffusion models by applying a fixed offset to the guidance embedding slice in cross-attention during inference.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.