Text-to-Image Generation
17개 벤치마크 · 논문 1,559편 · 이 태스크의 논문 보기 →
Benchmarks
COCO (Common Objects in Context)
GenEval
CUB
Multi-Modal-CelebA-HQ
DrawBench
Oxford 102 Flowers
Conceptual Captions
COCO
LHQC
MS-COCO
DPG
GeNeVA (CoDraw)
GeNeVA (i-CLEVR)
LAION COCO
T2I-CompBench
Colors
Flickr-8k
Most implemented
Show and Tell: A Neural Image Caption Generator
High-Resolution Image Synthesis with Latent Diffusion Models
Generative Adversarial Text to Image Synthesis
StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks
Papers
DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence
High-compression tokenizers are essential for scaling latent image generative models. However, aggressive compression creates a fundamental tradeoff between reconstruction fidelity and generation efficiency: high compres…
Text-to-Image GenerationImage ReconstructionHiRAE: Hierarchical Representation Autoencoding with Residual Budgets
Pretrained visual representations support image generation, but may not fully preserve the fine-grained details needed for faithful reconstruction. Meanwhile, intermediate encoder layers contain complementary visual deta…
Text-to-Image GenerationImproved Distributional Diffusion Models
Distributional Diffusion Models (DDMs) replace the standard mean-prediction denoiser with a distributional denoiser trained via a scoring rule objective, learning a stochastic approximation to p(x_1 mid x_t) rather than …
Text-to-Image GenerationUnifying Distributional Training for One-Step Visual Generation
Distributional training provides collective supervision for one-step visual generation by matching real and generated features in frozen representation spaces. We introduce a unified theoretical framework that separates …
Text-to-Image GenerationOn the Diffusibility of High-Dimensional Latents
Representation Autoencoders (RAEs) enable diffusion models to operate in the feature spaces of pretrained visual encoders. However, many off-the-shelf encoders are not optimized for faithful reconstruction, discarding fi…
Text-to-Image GenerationImage ReconstructionEfficient Text-to-Image Generation: An Adaptive Step Schedule Controller for Diffusion Models
Text-to-image diffusion models often use a fixed number of denoising steps, balancing time costs and image quality. However, the optimal number of steps depends on the complexity of the input text prompt. We propose an a…
Text-to-Image Generation