Daniel Cohen-Or

Fine-grained identity tuning enables precise facial edits in text-to-image models without additional training, preserving identity consistency across diverse outputs.

Localization signals in VLMs remain hidden in intermediate representations, leading to significant failures in current image editing pipelines.

Introducing a transport-based approach that ensures stable and expressive 3D stylizations by controlling style feature allocation across multiple views.

Semantic Browsing transforms image generation by allowing users to explore diverse visual interpretations through structured, meaningful variations rather than random noise.

Forget complex HDR-specific architectures: a simple log encoding and camera-aware training unlocks high-quality HDR video generation from off-the-shelf generative models.

Style transfer can now capture the essence of artistic abstraction, not just surface-level appearance, by explicitly reinterpreting image structure.

Unleashing creative potential in text-to-image models just got easier: on-the-fly repulsion in the contextual space lets you steer diffusion transformers towards richer diversity without sacrificing image quality or blowing your compute budget.

Steer diffusion models to seamlessly blend pasted objects into new contexts without prompts by selectively loosening positional encoding constraints based on saliency.

Achieve synchronized portrait video edits with Sync-LoRA, which propagates edits from a single frame while maintaining temporal coherence and identity consistency, even generalizing to unseen identities.

Forget tedious masks and prompts: EditP23 lets you edit 3D objects just by showing the model a before-and-after image pair.

Achieve semantically coherent image compositions by mixing layout-focused and appearance-focused visual representations in a diffusion model's cross-attention.