Anran Qi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Anran Qi has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Takeo Igarashi
University of Tokyo, Tokyo, Japan
10 shared publications - I.-Chao Shen
University of Tokyo, Tokyo, Japan
6 shared publications - Yuki Koyama
The University of Tokyo, Tokyo, Japan
4 shared publications - Yuki Tatsukawa
The University of Tokyo, Tokyo, Japan
4 shared publications - Adrien Bousseau
Centre Inria d'Université Côte d'Azur, Sophia Antipolis, France
3 shared publications - Ariel Shamir
Reichman University, Herzliya, Israel
3 shared publications
Latest Publications
- Open Access
- 30 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Many-Videos Browsing: Interactive Visualization of Generated Video Collections
- 30 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Open AccessMany-Videos Browsing: Interactive Visualization of Generated Video Collections
- Open Access
- 30 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Voronoi Scissors: Approximate Dissection of 2D Shapes with Continuous Optimization
- 30 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessVoronoi Scissors: Approximate Dissection of 2D Shapes with Continuous Optimization
- 13 November 2026
- Published by Wiley in Computer Graphics Forum
A Study of the Design Space of Motion and Disocclusion Control in Video Generation
Disocclusion occurs when object movement reveals previously hidden content. Existing image‐to‐video methods provide different forms of motion and disocclusion control, yet the design space they span remains poorly understood. In this work, we systematically explore an algorithmic design space spanning motion specification (text‐based versus tracking‐based) and dis‐occlusion specification (text‐based versus image‐based). To facilitate this study, we construct a benchmark consisting of both synthetic scenes and captured scenes, covering rigid and deformable motions, as well as static and dynamic disoccluded objects. We compare existing and adapted methods across this design space to analyze the strengths and limitations of different control strategies for handling disocclusions during motion. Our study reveals that tracking‐based guidance improves motion accuracy, while image‐based guidance improves the fidelity of newly revealed content. However, jointly controlling the appearance and dynamics of disoccluded objects remains a challenging open problem, particularly in scenarios where newly revealed objects undergo their own motion after becoming visible.
- 13 November 2026
- Published by Wiley in Computer Graphics Forum
A Study of the Design Space of Motion and Disocclusion Control in Video Generation
Disocclusion occurs when object movement reveals previously hidden content. Existing image‐to‐video methods provide different forms of motion and disocclusion control, yet the design space they span remains poorly understood. In this work, we systematically explore an algorithmic design space spanning motion specification (text‐based versus tracking‐based) and dis‐occlusion specification (text‐based versus image‐based). To facilitate this study, we construct a benchmark consisting of both synthetic scenes and captured scenes, covering rigid and deformable motions, as well as static and dynamic disoccluded objects. We compare existing and adapted methods across this design space to analyze the strengths and limitations of different control strategies for handling disocclusions during motion. Our study reveals that tracking‐based guidance improves motion accuracy, while image‐based guidance improves the fidelity of newly revealed content. However, jointly controlling the appearance and dynamics of disoccluded objects remains a challenging open problem, particularly in scenarios where newly revealed objects undergo their own motion after becoming visible.
- Open Access
- 27 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Rags2Riches: Computational Garment Reuse
- 27 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessRags2Riches: Computational Garment Reuse
- Open Access
- 27 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-11
GarmentImage: Raster Encoding of Garment Sewing Patterns with Diverse Topologies
- 27 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessGarmentImage: Raster Encoding of Garment Sewing Patterns with Diverse Topologies
- 27 May 2025
- Published by Elsevier in Computers & Graphics
- Vol. 130
Sketch2Data: Recovering data from hand-drawn infographics
- 27 May 2025
- Published by Elsevier in Computers & Graphics
- Vol. 130
Sketch2Data: Recovering data from hand-drawn infographics
- Open Access
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-14
FontCraft: Multimodal Font Design Using Interactive Bayesian Optimization
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-14
Open AccessFontCraft: Multimodal Font Design Using Interactive Bayesian Optimization
- Open Access
- 30 April 2024
- Published by Wiley in Computer Graphics Forum
- Vol. 43 (2)
FontCLIP: A Semantic Typography Visual‐Language Model for Multilingual Font Applications
Acquiring the desired font for various design tasks can be challenging and requires professional typographic knowledge. While previous font retrieval or generation works have alleviated some of these difficulties, they often lack support for multiple languages and semantic attributes beyond the training data domains. To solve this problem, we present FontCLIP – a model that connects the semantic understanding of a large vision‐language model with typographical knowledge. We integrate typography‐specific knowledge into the comprehensive vision‐language knowledge of a pretrained CLIP model through a novel finetuning approach. We propose to use a compound descriptive prompt that encapsulates adaptively sampled attributes from a font attribute dataset focusing on Roman alphabet characters. FontCLIP's semantic typographic latent space demonstrates two unprecedented generalization abilities. First, FontCLIP generalizes to different languages including Chinese, Japanese, and Korean (CJK), capturing the typographical features of fonts across different languages, even though it was only finetuned using fonts of Roman characters. Second, FontCLIP can recognize the semantic attributes that are not presented in the training data. FontCLIP's dual‐modality and generalization abilities enable multilingual and cross‐lingual font retrieval and letter shape optimization, reducing the burden of obtaining desired fonts.
- 30 April 2024
- Published by Wiley in Computer Graphics Forum
- Vol. 43 (2)
Open AccessFontCLIP: A Semantic Typography Visual‐Language Model for Multilingual Font Applications
Acquiring the desired font for various design tasks can be challenging and requires professional typographic knowledge. While previous font retrieval or generation works have alleviated some of these difficulties, they often lack support for multiple languages and semantic attributes beyond the training data domains. To solve this problem, we present FontCLIP – a model that connects the semantic understanding of a large vision‐language model with typographical knowledge. We integrate typography‐specific knowledge into the comprehensive vision‐language knowledge of a pretrained CLIP model through a novel finetuning approach. We propose to use a compound descriptive prompt that encapsulates adaptively sampled attributes from a font attribute dataset focusing on Roman alphabet characters. FontCLIP's semantic typographic latent space demonstrates two unprecedented generalization abilities. First, FontCLIP generalizes to different languages including Chinese, Japanese, and Korean (CJK), capturing the typographical features of fonts across different languages, even though it was only finetuned using fonts of Roman characters. Second, FontCLIP can recognize the semantic attributes that are not presented in the training data. FontCLIP's dual‐modality and generalization abilities enable multilingual and cross‐lingual font retrieval and letter shape optimization, reducing the burden of obtaining desired fonts.
- 18 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Graphics and Applications
- Vol. 44 (4) , 126-132
PerfectTailor: Scale-Preserving 2-D Pattern Adjustment Driven by 3-D Garment Editing
We address the problem of modifying a given well-designed 2D sewing pattern to accommodate garment edits in the 3D space. Existing methods usually adjust the sewing pattern by applying uniform flattening to the 3D garment. The problems are twofold: first, it ignores local scaling of the 2D sewing pattern such as shrinking ribs of cuffs; second, it does not respect the implicit design rules and conventions of the industry, such as the use of straight edges for simplicity and precision in sewing. To address those problems, we present a pattern adjustment method that considers the non-uniform local scaling of the 2D sewing pattern by utilizing the intrinsic scale matrix. In addition, we preserve the original boundary shape by an as-original-as-possible geometric constraint when desirable. We build a prototype with a set of commonly used alteration operations and showcase the capability of our method via a number of alteration examples throughout the paper.
- 18 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Graphics and Applications
- Vol. 44 (4) , 126-132
PerfectTailor: Scale-Preserving 2-D Pattern Adjustment Driven by 3-D Garment Editing
We address the problem of modifying a given well-designed 2D sewing pattern to accommodate garment edits in the 3D space. Existing methods usually adjust the sewing pattern by applying uniform flattening to the 3D garment. The problems are twofold: first, it ignores local scaling of the 2D sewing pattern such as shrinking ribs of cuffs; second, it does not respect the implicit design rules and conventions of the industry, such as the use of straight edges for simplicity and precision in sewing. To address those problems, we present a pattern adjustment method that considers the non-uniform local scaling of the 2D sewing pattern by utilizing the intrinsic scale matrix. In addition, we preserve the original boundary shape by an as-original-as-possible geometric constraint when desirable. We build a prototype with a set of commonly used alteration operations and showcase the capability of our method via a number of alteration examples throughout the paper.
- Open Access
FontCLIP: A Semantic Typography Visual-Language Model for Multilingual Font Applications
Acquiring the desired font for various design tasks can be challenging and requires professional typographic knowledge. While previous font retrieval or generation works have alleviated some of these difficulties, they often lack support for multiple languages and semantic attributes beyond the training data domains. To solve this problem, we present FontCLIP: a model that connects the semantic understanding of a large vision-language model with typographical knowledge. We integrate typography-specific knowledge into the comprehensive vision-language knowledge of a pretrained CLIP model through a novel finetuning approach. We propose to use a compound descriptive prompt that encapsulates adaptively sampled attributes from a font attribute dataset focusing on Roman alphabet characters. FontCLIP's semantic typographic latent space demonstrates two unprecedented generalization abilities. First, FontCLIP generalizes to different languages including Chinese, Japanese, and Korean (CJK), capturing the typographical features of fonts across different languages, even though it was only finetuned using fonts of Roman characters. Second, FontCLIP can recognize the semantic attributes that are not presented in the training data. FontCLIP's dual-modality and generalization abilities enable multilingual and cross-lingual font retrieval and letter shape optimization, reducing the burden of obtaining desired fonts.
Open AccessFontCLIP: A Semantic Typography Visual-Language Model for Multilingual Font Applications
Acquiring the desired font for various design tasks can be challenging and requires professional typographic knowledge. While previous font retrieval or generation works have alleviated some of these difficulties, they often lack support for multiple languages and semantic attributes beyond the training data domains. To solve this problem, we present FontCLIP: a model that connects the semantic understanding of a large vision-language model with typographical knowledge. We integrate typography-specific knowledge into the comprehensive vision-language knowledge of a pretrained CLIP model through a novel finetuning approach. We propose to use a compound descriptive prompt that encapsulates adaptively sampled attributes from a font attribute dataset focusing on Roman alphabet characters. FontCLIP's semantic typographic latent space demonstrates two unprecedented generalization abilities. First, FontCLIP generalizes to different languages including Chinese, Japanese, and Korean (CJK), capturing the typographical features of fonts across different languages, even though it was only finetuned using fonts of Roman characters. Second, FontCLIP can recognize the semantic attributes that are not presented in the training data. FontCLIP's dual-modality and generalization abilities enable multilingual and cross-lingual font retrieval and letter shape optimization, reducing the burden of obtaining desired fonts.