We introduce a novel diffusion transformer, LazyDiffusion, that generates partial image updates efficiently. Our approach targets interactive image editing applications in which, starting from a blank canvas or an image, a user specifies a sequence of localized image modifications using binary masks and text prompts. Our generator operates in two phases. First, a context encoder processes the current canvas and user mask to produce a compact global context tailored to the region to generate. Second, conditioned on this context, a diffusion-based transformer decoder synthesizes the masked pixels in a "lazy" fashion, i.e., it only generates the masked region. This contrasts with previous works that either regenerate the full canvas, wasting time and computation, or confine processing to a tight rectangular crop around the mask, ignoring the global image context altogether. Our decoder's runtime scales with the mask size, which is typically small, while our encoder introduces negligible overhead. We demonstrate that our approach is competitive with state-of-the-art inpainting methods in terms of quality and fidelity while providing a 10x speedup for typical user interactions, where the editing mask represents 10% of the image.
Microsoft Dynamics 365 AI Research（微软 Dynamics 365 人工智能研究院）；Duke University（杜克大学）；Microsoft Research（微软研究院）
Most existing text-to-image synthesis tasks are static single-turn generation, based on pre-defined textual descriptions of images. To explore more practical and interactive real-life applications, we introduce a new task - Interactive Image Editing, where users can guide an agent to edit images via multi-turn textual commands on-the-fly. In each session, the agent takes a natural language description from the user as the input and modifies the image generated in the previous turn to a new design, following the user description. The main challenges in this sequential and interactive image gene…
Generative adversarial networks (GANs) have enabled photorealistic image synthesis and editing. However, due to the high computational cost of large-scale generators (e.g., StyleGAN2), it usually takes seconds to see the results of a single edit on edge devices, prohibiting interactive user experience. In this paper, we take inspirations from modern rendering software and propose Anycost GAN for interactive natural image editing. We train the Anycost GAN to support elastic resolutions and channels for faster image generation at versatile speeds. Running subsets of the full generator produce ou…
Tsinghua University（清华大学）；Tencent AI Lab（腾讯人工智能实验室）；ByteDance Inc.（字节跳动公司）
Existing 3D-aware facial generation methods face a dilemma in quality versus editability: they either generate editable results in low resolution or high-quality ones with no editing flexibility. In this work, we propose a new approach that brings the best of both worlds together. Our system consists of three major components: (1) a 3D-semantics-aware generative model that produces view-consistent, disentangled face images and semantic masks; (2) a hybrid GAN inversion approach that initialize the latent codes from the semantic and texture encoder, and further optimized them for faithful recon…
SceneCraft: Interactive System for Image Editing via Scene Graph
University of Science, Ho Chi Minh, Vietnam（胡志明市理科大学）；Vietnam National University, Ho Chi Minh, Vietnam（越南国家大学胡志明市）；University of Dayton, Dayton, Ohio, USA（代顿大学）
Pay Attention and Move Better: Harnessing Attention for Interactive Motion Generation and Training-free Editing
Tsinghua University（清华大学）；The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））；The University of Hong Kong（香港大学）；Shanghai AI Laboratory（上海人工智能实验室）
Adobe Research（奥多比研究院）；Australian National University（澳大利亚国立大学）
Controllable image synthesis with user scribbles is a topic of keen interest in the computer vision community. In this paper, for the first time we study the problem of photorealistic image synthesis from incomplete and primitive human paintings. In particular, we propose a novel approach paint2pix, which learns to predict (and adapt) "what a user wants to draw" from rudimentary brushstroke inputs, by learning a mapping from the manifold of incomplete human paintings to their realistic renderings. When used in conjunction with recent works in autonomous painting agents, we show that paint2pix …
I2E: From Image Pixels to Actionable Interactive Environments for Text-Guided Image Editing
Huazhong University of Science and Technology（华中科技大学）；Kuaishou Technology（快手科技）；Central China Normal University（华中师范大学）；Tsinghua University（清华大学）；Shanghai AI Laboratory（上海人工智能实验室）
University of Bern（伯尔尼大学）；University of Maryland, College Park（马里兰大学帕克分校）
We present a novel system for sketch-based face image editing, enabling users to edit images intuitively by sketching a few strokes on a region of interest. Our interface features tools to express a desired image manipulation by providing both geometry and color constraints as user-drawn strokes. As an alternative to the direct user input, our proposed system naturally supports a copy-paste mode, which allows users to edit a given image region by using parts of another exemplar image without the need of hand-drawn sketching at all. The proposed interface runs in real-time and facilitates an in…
FOCUS: Unified Vision-Language Modeling for Interactive Editing Driven by Referential Segmentation
Foundation Model Research Center, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所基础模型研究中心）；School of Artificial Intelligence, University of Chinese Academy of Science（中国科学院大学人工智能学院）；Peng Cheng Laboratory, Shenzhen, China（鹏城实验室）；Wuhan AI Research, Wuhan, China（武汉人工智能研究所）
FED-Bench: A Cross-Granular Benchmark for Disentangled Evaluation of Facial Expression Editing
Xi’an Jiaotong University（西安交通大学）；Northwestern Polytechnical University（西北工业大学）；National University of Singapore（新加坡国立大学）；Fudan University（复旦大学）；Zhejiang Lab（之江实验室）
NVIDIA（英伟达）；University of Toronto（多伦多大学）；Vector Institute（矢量研究所）；MIT（麻省理工学院）
Generative adversarial networks (GANs) have recently found applications in image editing. However, most GAN based image editing methods often require large scale datasets with semantic segmentation annotations for training, only provide high level control, or merely interpolate between different images. Here, we propose EditGAN, a novel method for high quality, high precision semantic image editing, allowing users to edit images by modifying their highly detailed part segmentation masks, e.g., drawing a new mask for the headlight of a car. EditGAN builds on a GAN framework that jointly models …
ArtisanGS: Interactive Tools for Gaussian Splat Selection with AI and Human in the Loop
NVIDIA；NVIDIA Canada（NVIDIA 加拿大）；University of Toronto Canada（多伦多大学 加拿大）；NVIDIA France（NVIDIA 法国）；University of Toronto（多伦多大学）；Institute for Clarity in Documentation Dublin Ohio USA（文档清晰研究所 俄亥俄州 大致）；Inria Paris-Rocquencourt（法国巴黎-罗克琴特研究所）；Rajiv Gandhi University Doimukh Arunachal Pradesh India（拉吉夫·甘地大学 亚当穆克 阿鲁纳恰尔邦 印度）；Tsinghua University Haidian Qu Beijing Shi China（清华大学 海淀区 北京市 中国）；Palmer Research Laboratories San Antonio Texas USA（帕勒研究中心 肯塔基州 威斯康星州 美国）
University of Science and Technology of China（中国科学技术大学）；Snap Inc（斯奈普公司）；Microsoft Cloud AI（微软云人工智能）；City University of Hong Kong（香港城市大学）
Despite the recent success of face image generation with GANs, conditional hair editing remains challenging due to the under-explored complexity of its geometry and appearance. In this paper, we present MichiGAN (Multi-Input-Conditioned Hair Image GAN), a novel conditional image generation method for interactive portrait hair manipulation. To provide user control over every major hair visual factor, we explicitly disentangle hair into four orthogonal attributes, including shape, structure, appearance, and background. For each of them, we design a corresponding condition module to represent, pr…
University of California, Los Angeles（加利福尼亚大学洛杉矶分校）
Significant progress has been made by the advances in Generative Adversarial Networks (GANs) for image generation. However, there lacks enough understanding of how a realistic image is generated by the deep representations of GANs from a random vector. This chapter gives a summary of recent works on interpreting deep generative models. The methods are categorized into the supervised, the unsupervised, and the embedding-guided approaches. We will see how the human-understandable concepts that emerge in the learned representation can be identified and used for interactive image generation and ed…
DPDEdit: Detail-Preserved Diffusion Models for Multimodal Fashion Image Editing
Shenzhen Technology University（深圳科技大学）；University of Washington（华盛顿大学）；Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences（深圳先进技术研究院，中国科学院）
Generative Adversarial Network (GAN) based localized image editing can suffer from ambiguity between semantic attributes. We thus present a novel objective function to evaluate the locality of an image edit. By introducing the supervision from a pre-trained segmentation network and optimizing the objective function, our framework, called Locally Effective Latent Space Direction (LELSD), is applicable to any dataset and GAN architecture. Our method is also computationally fast and exhibits a high extent of disentanglement, which allows users to interactively perform a sequence of edits on an im…
Beijing Academy of Artificial Intelligence（北京人工智能研究院）；University of Chinese Academy of Sciences（中国科学院大学）；Shenyang Institute of Computing Technology, Chinese Academy of Sciences（中国科学院沈阳计算技术研究所）；Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）
Inria, Mimesis Group（法国国家信息与自动化研究所，Mimesis 团队）；CNRS, University of Strasbourg（法国国家科学研究中心，斯特拉斯堡大学）；Technical University of Munich（慕尼黑工业大学）
We present Calipso, an interactive method for editing images and videos in a physically-coherent manner. Our main idea is to realize physics-based manipulations by running a full physics simulation on proxy geometries given by non-rigidly aligned CAD models. Running these simulations allows us to apply new, unseen forces to move or deform selected objects, change physical parameters such as mass or elasticity, or even add entire new objects that interact with the rest of the underlying scene. In Calipso, the user makes edits directly in 3D; these edits are processed by the simulation and then …
Latent space exploration is a technique that discovers interpretable latent directions and manipulates latent codes to edit various attributes in images generated by generative adversarial networks (GANs). However, in previous work, spatial control is limited to simple transformations (e.g., translation and rotation), and it is laborious to identify appropriate latent directions and adjust their parameters. In this paper, we tackle the problem of editing the StyleGAN image layout by annotating the image directly. To do so, we propose an interactive framework for manipulating latent codes in ac…
Trinity College Dublin（都柏林圣三一学院）；Adobe Research（Adobe研究）
Interactive object cutout tools are the cornerstone of the image editing workflow. Recent deep-learning based interactive segmentation algorithms have made significant progress in handling complex images and rough binary selections can typically be obtained with just a few clicks. Yet, deep learning techniques tend to plateau once this rough selection has been reached. In this work, we interpret this plateau as the inability of current algorithms to sufficiently leverage each user interaction and also as the limitations of current training/testing datasets. We propose a novel interactive archi…
