Chunpeng Li

Chunpeng Li

Auto-generated profile, based on publicly available information

Chunpeng Li

Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China

Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 June 2020 in Computational Visual Media
Verified
Journal Article
Open access
PDF file
YXYun-Peng Xiao
Yun-Peng Xiao
YLYu-Kun Lai
Yu-Kun Lai
FZFanglue Zhang
Fanglue Zhang
CLChunpeng Li
Chunpeng Li
LGLin Gao
Lin Gao
Researchers have achieved great success in dealing with 2D images using deep learning. In recent years, 3D computer vision and geometry deep learning have gained ever more attention. Many advanced techniques for 3D shapes have been proposed for different applications. Unlike 2D images, which can be uniformly represented by a regular grid of pixels, 3D shapes have various representations, such as depth images, multi-view images, voxels, point clouds, meshes, implicit surfaces, etc. The performance achieved in different applications largely depends on the representation used, and there is no unique representation that works well for all applications. Therefore, in this survey, we review recent developments in deep learning for 3D geometry from a representation perspective, summarizing the advantages and disadvantages of different representations for different applications. We also present existing datasets in these representations and further discuss future research directions.

26 July 2023 in ACM Transactions on Graphics
Verified
Journal Article
Open access
LGLin Gao
Lin Gao
FLFeng-Lin Liu
Feng-Lin Liu
SCShu-Yu Chen
Shu-Yu Chen
KJKaiwen Jiang
Kaiwen Jiang
CLChun-Peng Li
Chun-Peng Li
HFHongbo Fu
Hongbo Fu
Realistic 3D facial generation based on Neural Radiance Fields (NeRFs) from 2D sketches benefits various applications. Despite the high realism of free-view rendering results of NeRFs, it is tedious and difficult for artists to achieve detailed 3D control and manipulation. Meanwhile, due to its conciseness and expressiveness, sketching has been widely used for 2D facial image generation and editing. Applying sketching to NeRFs is challenging due to the inherent uncertainty for 3D generation with 2D constraints, a significant gap in content richness when generating faces from sparse sketches, and potential inconsistencies for sequential multi-view editing given only 2D sketch inputs. To address these challenges, we present SketchFaceNeRF, a novel sketch-based 3D facial NeRF generation and editing method, to produce free-view photo-realistic images. To solve the challenge of sketch sparsity, we introduce a Sketch Tri-plane Prediction net to first inject the appearance into sketches, thus generating features given reference images to allow color and texture control. Such features are then lifted into compact 3D tri-planes to supplement the absent 3D information, which is important for improving robustness and faithfulness. However, during editing, consistency for unseen or unedited 3D regions is difficult to maintain due to limited spatial hints in sketches. We thus adopt a Mask Fusion module to transform free-view 2D masks (inferred from sketch editing operations) into the tri-plane space as 3D masks, which guide the fusion of the original and sketch-based generated faces to synthesize edited faces. We further design an optimization approach with a novel space loss to improve identity retention and editing faithfulness. Our pipeline enables users to flexibly manipulate faces from different viewpoints in 3D space, easily designing desirable facial models. Extensive experiments validate that our approach is superior to the state-of-the-art 2D sketch-based image generation and editing approaches in realism and faithfulness.

19 July 2021 in ACM Transactions on Graphics
Verified
Journal Article
Open access
SCShu-Yu Chen
Shu-Yu Chen
FLFeng-Lin Liu
Feng-Lin Liu
YLYu-Kun Lai
Yu-Kun Lai
Paul L. RosinPaul L. Rosin
Paul L. Rosin
CLChunpeng Li
Chunpeng Li
LGLin Gao
Lin Gao
Recent facial image synthesis methods have been mainly based on conditional generative models. Sketch-based conditions can effectively describe the geometry of faces, including the contours of facial components, hair structures, as well as salient edges (e.g., wrinkles) on face surfaces but lack effective control of appearance, which is influenced by color, material, lighting condition, etc. To have more control of generated results, one possible approach is to apply existing disentangling works to disentangle face images into geometry and appearance representations. However, existing disentangling methods are not optimized for human face editing, and cannot achieve fine control of facial details such as wrinkles. To address this issue, we propose DeepFaceEditing, a structured disentanglement framework specifically designed for face images to support face generation and editing with disentangled control of geometry and appearance. We adopt a local-to-global approach to incorporate the face domain knowledge: local component images are decomposed into geometry and appearance representations, which are fused consistently using a global fusion module to improve generation quality. We exploit sketches to assist in extracting a better geometry representation, which also supports intuitive geometry editing via sketching. The resulting method can either extract the geometry and appearance representations from face images, or directly extract the geometry representation from face sketches. Such representations allow users to easily edit and synthesize face images, with decoupled control of their geometry and appearance. Both qualitative and quantitative evaluations show the superior detail and appearance control abilities of our method compared to state-of-the-art methods.

28 February 2024 in IEEE Robotics and Automation Letters
Verified
Journal Article
Closed access
ZYZejun Yang
Zejun Yang
LNLi Ning
Li Ning
HWHaitao Wang
Haitao Wang
TJTianyu Jiang
Tianyu Jiang
SZShaolin Zhang
Shaolin Zhang
ZWZhaoqi Wang
Zhaoqi Wang
To complete tasks in dynamic environments, robots need to timely update their plans to react to environment changes. Traditional stripe-like or learning-based planners struggle to achieve this due to their high reliance on meticulously predefined planning rules or labeled data. Fortunately, recent works find that Large Language Models (LLMs) can be effectively prompted to solve planning problems. Thus, we investigate the strategies for LLMs to master reactive planning problems without complex definitions and extra training. We propose Text2Reaction, an LLM-based framework enabling robots to continuously reason and update plans according to the latest environment changes. Inspired from human's step-by-step re-planning process, we present the Re-planning Prompt, which informs LLMs the basic principles of re-planning and fosters the gradual development of a current plan to a new one in a three-hop reasoning manner–cause analysis, consequence inference, and plan adjustment. In addition, Text2Reaction is designed to first generate an initial plan based on the task description before execution, allowing for subsequent iterative updates of this plan. We demonstrate the superior performance of Text2Reaction over prior works in reacting to various environment changes and completing varied tasks. In addition, we validate the reliability of our re-planning prompt through ablation experiments and its capability when deployed in real-world robots, enabling continuous reasoning in the face of diverse changes until the user instructions are successfully completed.

21 August 2015 in Sensors
Verified
Journal Article
Open access
PDF file
JLJing Liu
Jing Liu
CLChunpeng Li
Chunpeng Li
XFXuefeng Fan
Xuefeng Fan
ZWZhaoqi Wang
Zhaoqi Wang
Depth estimation is a classical problem in computer vision, which typically relies on either a depth sensor or stereo matching alone. The depth sensor provides real-time estimates in repetitive and textureless regions where stereo matching is not effective. However, stereo matching can obtain more accurate results in rich texture regions and object boundaries where the depth sensor often fails. We fuse stereo matching and the depth sensor using their complementary characteristics to improve the depth estimation. Here, texture information is incorporated as a constraint to restrict the pixel’s scope of potential disparities and to reduce noise in repetitive and textureless regions. Furthermore, a novel pseudo-two-layer model is used to represent the relationship between disparities in different pixels and segments. It is more robust to luminance variation by treating information obtained from a depth sensor as prior knowledge. Segmentation is viewed as a soft constraint to reduce ambiguities caused by under- or over-segmentation. Compared to the average error rate 3.27% of the previous state-of-the-art methods, our method provides an average error rate of 2.61% on the Middlebury datasets, which shows that our method performs almost 20% better than other “fused” algorithms in the aspect of precision.

1 November 2010
Verified
Conference Paper
Closed access
DZDan Zong
Dan Zong
CLChunpeng Li
Chunpeng Li
SXShihong Xia
Shihong Xia
ZWZhaoqi Wang
Zhaoqi Wang
In recent years, virtual assembly has become an effective technique for large equipment's maintainability analysis and verification. As an indispensable part in maintenance process simulation, motion control for virtual human still remains on low level based on key-frame or inverse kinematics, which in turn leads to an over complicated modeling process. Therefore, we propose an instruction-level motion control method of virtual human in maintenance process simulation in this paper, which receives the maintenance instruction as input and converts it into natural looking animation sequence in order to simulate maintenance process rapidly. There are two layers in our method: primitive layer and script parsing layer. The primitive layer contains a series of parameterized models, which are called primitives. Each primitive generates a specific motion sequence given a set of intuitive parameters. The script parsing layer converts the instructions into several kinds of primitives by using modules. Each maintenance instruction can be converted into a set of modules and then into primitives. Our method provides a high-level motion control interface and generates natural looking motions interactively. Experiments have proven the effectiveness.