InsHuman: Towards Natural and Identity-Preserving Human Insertion

• URL: http://arxiv.org/abs/2605.07402v1
• Date: Fri, 08 May 2026 07:58:03 GMT
• Title: InsHuman: Towards Natural and Identity-Preserving Human Insertion
• Abstract summary: We propose InsHuman for natural and identity-preserving human insertion.<n> Specifically, we propose Human-Background Adaptive Fusion (HBAF), which detects foreground humans to obtain a binary mask.<n>We also propose Face-to-Face ID-Preserving (FFIP), which detects and matches faces between the generated image and the source image.
• Score: 37.70158864105303
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Human insertion aims to naturally place specific individuals into a target background. Although existing image editing models may have such ability, they often produce failure cases, including inappropriate human pose in new background, inconsistent number of people, and modified facial identity. Moreover, publicly available human datasets often lack full-body portraits and realistic physical interaction between humans and their background. To address these challenges, we propose InsHuman for natural and identity-preserving human insertion. Specifically, we propose Human-Background Adaptive Fusion (HBAF), which detects foreground humans to obtain a binary mask and applies region-aware weighting to align the human regions between predicted and ground-truth latents, ensuring the person's pose, count, and overall appearance are coherently adapted to the target background.We further propose Face-to-Face ID-Preserving (FFIP), which detects and matches faces between the generated image and the source image in terms of face recognition features to enforce identity consistency for each face.In addition, we propose Bidirectional Data Pairing (BDP) strategy to construct BDP-InsHuman, a high-quality dataset with realistic human-background interactions. Experiments demonstrate that InsHuman achieves significant improvements in generating plausible images while keeping human identity unchanged.

Related papers

• HOComp: Interaction-Aware Human-Object Composition [62.93211305213214]
  HOComp is a novel approach for compositing a foreground object onto a human-centric background image.<n> Experimental results on our dataset show that HOComp effectively generates human-object interactions with consistent appearances.
  arXiv  Detail & Related papers  (2025-07-22T17:59:21Z)
• Improving Personalized Image Generation through Social Context Feedback [15.582260415127935]
  We propose to overcome these shortcomings through feedback-based fine-tuning of existing personalized generation methods.<n>State-of-art detectors of pose, human-object-interaction, human facial recognition and human gaze-point estimation are used to refine the diffusion model.<n>The images generated in this manner show an improvement in the generated interactions, facial identities and image quality over three benchmark datasets.
  arXiv  Detail & Related papers  (2025-07-21T22:36:30Z)
• Leveraging Generative AI Models to Explore Human Identity [1.8139774424530175]
  This paper attempts to explore human identity by utilizing neural networks in an indirect manner.<n>By relating the generated human face to human identity, we establish a correspondence between the face image generation process and the process of human identity formation.
  arXiv  Detail & Related papers  (2025-04-19T12:35:07Z)
• Turn That Frown Upside Down: FaceID Customization via Cross-Training Data [49.51940625552275]
  CrossFaceID is the first large-scale, high-quality, and publicly available dataset designed to improve the facial modification capabilities of FaceID customization models.<n>It consists of 40,000 text-image pairs from approximately 2,000 persons, with each person represented by around 20 images showcasing diverse facial attributes.<n>During the training stage, a specific face of a person is used as input, and the FaceID customization model is forced to generate another image of the same person but with altered facial features.<n>Experiments show that models fine-tuned on the CrossFaceID dataset its performance in preserving FaceID fidelity while significantly improving its
  arXiv  Detail & Related papers  (2025-01-26T05:27:38Z)
• PersonaHOI: Effortlessly Improving Personalized Face with Human-Object Interaction Generation [62.85437763951757]
  PersonaHOI is a training- and tuning-free framework that fuses a general StableDiffusion model with a personalized face diffusion (PFD) model to generate identity-consistent human-object interaction (HOI) images.
  arXiv  Detail & Related papers  (2025-01-10T10:01:36Z)
• Boost Your Human Image Generation Model via Direct Preference Optimization [5.9726297901501475]
  Human image generation is a key focus in image synthesis due to its broad applications, but even slight inaccuracies in anatomy, pose, or details can compromise realism.<n>We explore Direct Preference Optimization (DPO), which trains models to generate preferred (winning) images while diverging from non-preferred (losing) ones.<n>We propose an enhanced DPO approach that incorporates high-quality real images as winning images, encouraging outputs to resemble real images rather than generated ones.
  arXiv  Detail & Related papers  (2024-05-30T16:18:05Z)
• StableIdentity: Inserting Anybody into Anywhere at First Sight [57.99693188913382]
  We propose StableIdentity, which allows identity-consistent recontextualization with just one face image. We are the first to directly inject the identity learned from a single image into video/3D generation without finetuning.
  arXiv  Detail & Related papers  (2024-01-29T09:06:15Z)
• Graph-based Generative Face Anonymisation with Pose Preservation [49.18049578591058]
  AnonyGAN is a GAN-based solution for face anonymisation. It replaces the visual information corresponding to a source identity with a condition identity provided as any single image.
  arXiv  Detail & Related papers  (2021-12-10T12:58:17Z)
• FusiformNet: Extracting Discriminative Facial Features on Different Levels [0.0]
  I propose FusiformNet, a novel framework for feature extraction that leverages the nature of discriminative facial features. FusiformNet achieved a state-of-the-art accuracy of 96.67% without labeled outside data, image augmentation, normalization, or special loss functions. Considering its ability to extract both general and local facial features, the utility of FusiformNet may not be limited to facial recognition but also extend to other DNN-based tasks.
  arXiv  Detail & Related papers  (2020-11-01T18:00:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.