SSR-Encoder: Encoding Selective Subject Representation for Subject-Driven Generation

TL;DR

The SSR-Encoder is a new architecture that effectively captures and encodes subject representations from images and text, enabling precise and versatile subject-driven image generation without test-time fine-tuning.

Contribution

It introduces the SSR-Encoder with a Token-to-Patch Aligner and Detail-Preserving Subject Encoder, advancing subject representation and control in image generation.

Findings

Effective in zero-shot subject-driven generation

Supports multiple query modalities including text and masks

Demonstrates broad applicability and high quality in experiments

Abstract

Recent advancements in subject-driven image generation have led to zero-shot generation, yet precise selection and focus on crucial subject representations remain challenging. Addressing this, we introduce the SSR-Encoder, a novel architecture designed for selectively capturing any subject from single or multiple reference images. It responds to various query modalities including text and masks, without necessitating test-time fine-tuning. The SSR-Encoder combines a Token-to-Patch Aligner that aligns query inputs with image patches and a Detail-Preserving Subject Encoder for extracting and preserving fine features of the subjects, thereby generating subject embeddings. These embeddings, used in conjunction with original text embeddings, condition the generation process. Characterized by its model generalizability and efficiency, the SSR-Encoder adapts to a range of custom models and control modules. Enhanced by the Embedding Consistency Regularization Loss for improved training, our extensive experiments demonstrate its effectiveness in versatile and high-quality image generation, indicating its broad applicability. Project page: https://ssr-encoder.github.io