Search

Creating photorealistic avatars for individuals traditionally involves extensive capture sessions with complex and expensive studio devices like the LightStage system. While recent strides in neural representations have enabled the generation of photorealistic and animatable 3D avatars from quick phone scans, they have the capture-time lighting baked-in, lack facial details and have missing regions in areas such as the back of the ears. Thus, they lag in quality compared to studio-captured avatars. In this paper, we propose a method that bridges this gap by generating studio-like illuminated texture maps from short, monocular phone captures. We do this by parameterizing the phone texture maps using the $W^+$ space of a StyleGAN2, enabling near-perfect reconstruction. Then, we finetune a StyleGAN2 by sampling in the $W^+$ parameterized space using a very small set of studio-captured textures as an adversarial training signal. To further enhance the realism and accuracy of facial details, we super-resolve the output of the StyleGAN2 using carefully designed diffusion model that is guided by image gradients of the phone-captured texture map. Once trained, our method excels at producing studio-like facial texture maps from casual monocular smartphone videos. Demonstrating its capabilities, we showcase the generation of photorealistic, uniformly lit, complete avatars from monocular phone captures. The project page can be found at http://shahrukhathar.github.io/2024/07/22/Bridging.html, Comment: ECCV 2024

Faithful real-time facial animation is essential for avatar-mediated telepresence in Virtual Reality (VR). To emulate authentic communication, avatar animation needs to be efficient and accurate: able to capture both extreme and subtle expressions within a few milliseconds to sustain the rhythm of natural conversations. The oblique and incomplete views of the face, variability in the donning of headsets, and illumination variation due to the environment are some of the unique challenges in generalization to unseen faces. In this paper, we present a method that can animate a photorealistic avatar in realtime from head-mounted cameras (HMCs) on a consumer VR headset. We present a self-supervised learning approach, based on a cross-view reconstruction objective, that enables generalization to unseen users. We present a lightweight expression calibration mechanism that increases accuracy with minimal additional cost to run-time efficiency. We present an improved parameterization for precise ground-truth generation that provides robustness to environmental variation. The resulting system produces accurate facial animation for unseen users wearing VR headsets in realtime. We compare our approach to prior face-encoding methods demonstrating significant improvements in both quantitative metrics and qualitative results., Comment: SIGGRAPH 2024 (ACM Transactions on Graphics (TOG))

Existing photorealistic relightable hand models require extensive identity-specific observations in different views, poses, and illuminations, and face challenges in generalizing to natural illuminations and novel identities. To bridge this gap, we present URHand, the first universal relightable hand model that generalizes across viewpoints, poses, illuminations, and identities. Our model allows few-shot personalization using images captured with a mobile phone, and is ready to be photorealistically rendered under novel illuminations. To simplify the personalization process while retaining photorealism, we build a powerful universal relightable prior based on neural relighting from multi-view images of hands captured in a light stage with hundreds of identities. The key challenge is scaling the cross-identity training while maintaining personalized fidelity and sharp details without compromising generalization under natural illuminations. To this end, we propose a spatially varying linear lighting model as the neural renderer that takes physics-inspired shading as input feature. By removing non-linear activations and bias, our specifically designed lighting model explicitly keeps the linearity of light transport. This enables single-stage training from light-stage data while generalizing to real-time rendering under arbitrary continuous illuminations across diverse identities. In addition, we introduce the joint learning of a physically based model and our neural relighting model, which further improves fidelity and generalization. Extensive experiments show that our approach achieves superior performance over existing methods in terms of both quality and generalizability. We also demonstrate quick personalization of URHand from a short phone scan of an unseen identity., Comment: Project Page https://frozenburning.github.io/projects/urhand/

Hair plays a significant role in personal identity and appearance, making it an essential component of high-quality, photorealistic avatars. Existing approaches either focus on modeling the facial region only or rely on personalized models, limiting their generalizability and scalability. In this paper, we present a novel method for creating high-fidelity avatars with diverse hairstyles. Our method leverages the local similarity across different hairstyles and learns a universal hair appearance prior from multi-view captures of hundreds of people. This prior model takes 3D-aligned features as input and generates dense radiance fields conditioned on a sparse point cloud with color. As our model splits different hairstyles into local primitives and builds prior at that level, it is capable of handling various hair topologies. Through experiments, we demonstrate that our model captures a diverse range of hairstyles and generalizes well to challenging new hairstyles. Empirical results show that our method improves the state-of-the-art approaches in capturing and generating photorealistic, personalized avatars with complete hair.

This study explores the integration of generative artificial intelligence (AI), specifically large language models, with multi-modal analogical reasoning as an innovative approach to enhance science, technology, engineering, and mathematics (STEM) education. We have developed a novel system that utilizes the capacities of generative AI to transform intricate principles in mathematics, physics, and programming into comprehensible metaphors. To further augment the educational experience, these metaphors are subsequently converted into visual form. Our study aims to enhance the learners' understanding of STEM concepts and their learning engagement by using the visual metaphors. We examine the efficacy of our system via a randomized A/B/C test, assessing learning gains and motivation shifts among the learners. Our study demonstrates the potential of applying large language models to educational practice on STEM subjects. The results will shed light on the design of educational system in terms of harnessing AI's potential to empower educational stakeholders.

The Large Language Models (LLMs), such as GPT and BERT, were proposed for natural language processing (NLP) and have shown promising results as general-purpose language models. An increasing number of industry professionals and researchers are adopting LLMs for program analysis tasks. However, one significant difference between programming languages and natural languages is that a programmer has the flexibility to assign any names to variables, methods, and functions in the program, whereas a natural language writer does not. Intuitively, the quality of naming in a program affects the performance of LLMs in program analysis tasks. This paper investigates how naming affects LLMs on code analysis tasks. Specifically, we create a set of datasets with code containing nonsense or misleading names for variables, methods, and functions, respectively. We then use well-trained models (CodeBERT) to perform code analysis tasks on these datasets. The experimental results show that naming has a significant impact on the performance of code analysis tasks based on LLMs, indicating that code representation learning based on LLMs heavily relies on well-defined names in code. Additionally, we conduct a case study on some special code analysis tasks using GPT, providing further insights., Comment: 3 Table, 8 figures

Programming skills are rapidly becoming essential for many educational paths and career opportunities. Yet, for many international students, the traditional approach to teaching introductory programming courses can be a significant challenge due to the complexities of the language, the lack of prior programming knowledge, and the language and cultural barriers. This study explores how large language models and gamification can scaffold coding learning and increase Chinese students sense of belonging in introductory programming courses. In this project, a gamification intelligent tutoring system was developed to adapt to Chinese international students learning needs and provides scaffolding to support their success in introductory computer programming courses., Comment: Doctoral consortium for IUI 2023

The Bidirectional Encoder Representations from Transformers (BERT) were proposed in the natural language process (NLP) and shows promising results. Recently researchers applied the BERT to source-code representation learning and reported some good news on several downstream tasks. However, in this paper, we illustrated that current methods cannot effectively understand the logic of source codes. The representation of source code heavily relies on the programmer-defined variable and function names. We design and implement a set of experiments to demonstrate our conjecture and provide some insights for future works., Comment: 1 table, 2 figures

The capture and animation of human hair are two of the major challenges in the creation of realistic avatars for the virtual reality. Both problems are highly challenging, because hair has complex geometry and appearance, as well as exhibits challenging motion. In this paper, we present a two-stage approach that models hair independently from the head to address these challenges in a data-driven manner. The first stage, state compression, learns a low-dimensional latent space of 3D hair states containing motion and appearance, via a novel autoencoder-as-a-tracker strategy. To better disentangle the hair and head in appearance learning, we employ multi-view hair segmentation masks in combination with a differentiable volumetric renderer. The second stage learns a novel hair dynamics model that performs temporal hair transfer based on the discovered latent codes. To enforce higher stability while driving our dynamics model, we employ the 3D point-cloud autoencoder from the compression stage for de-noising of the hair state. Our model outperforms the state of the art in novel view synthesis and is capable of creating novel hair animations without having to rely on hair observations as a driving signal. Project page is here https://ziyanw1.github.io/neuwigs/.

This paper investigates a novel method for designing linear precoders with finite alphabet inputs based on autoencoders (AE) without the knowledge of the channel model. By model-free training of the autoencoder in a multiple-input multiple-output (MIMO) system, the proposed method can effectively solve the optimization problem to design the precoders that maximize the mutual information between the channel inputs and outputs, when only the input-output information of the channel can be observed. Specifically, the proposed method regards the receiver and the precoder as two independent parameterized functions in the AE and alternately trains them using the exact and approximated gradient, respectively. Compared with previous precoders design methods, it alleviates the limitation of requiring the explicit channel model to be known. Simulation results show that the proposed method works as well as those methods under known channel models in terms of maximizing the mutual information and reducing the bit error rate., Comment: Accepted by GLOBECOM 2022

We present a study of the bulge-to-total ratio (B/T) of a Ks-band-selected sample of 88 close major-merger pairs of galaxies (H-KPAIR) based on 2-D decomposition of SDSS r-band images with \textsc{galfit}. We investigate the dependence of the interaction-induced specific star formation rate enhancement ($\rm sSFR_{enh}$) on the B/T ratio, and the effects of this dependence on the differences between star-forming galaxies (SFGs) in spiral+spiral (S+S) and spiral+elliptical (S+E) pairs. Of all 132 spiral galaxies in H-KPAIR, the 44 in S+E pairs show higher B/T than those in the 44 S+S pairs, with means of $\rm B/T = 0.35 \pm 0.05$ and $\rm B/T = 0.26 \pm 0.03$, respectively. There is a strong negative dependence of $\rm sSFR_{enh}$ on the B/T ratio and only paired SFGs with $\rm B/T<0.3$ show significant ($>5\sigma$) enhancement. Paired SFGs in S+S pairs show a similar trend, and many disky SFGs ($\rm B/T<0.1$) in S+S have strong sSFR enhancements ($\rm sSFR_{enh} > 0.7$~dex). For SFGs in S+E, the sSFR has no clear B/T dependence, nor any significant enhancement in any B/T bin. Disky SFGs in S+S show significant ($>4\sigma$) enhancement in the molecular gas content ($\rm M_{H_2}/M_{star}$), while SFGs in S+E have no such enhancement in any B/T bin. No significant enhancement on total gas content ($\rm M_{gas}/M_{star}$) is found in any B/T bin for paired galaxies. The star formation efficiency of either the total gas ($\rm SFE_{gas} = SFR/M_{gas}$) or the molecular gas ($\rm SFE_{H_2} = SFR/M_{H_2}$) does not depend on the B/T ratio. The only significant ($>4\sigma$) SFE enhancement found for paired SFGs is the $\rm SFE_{gas}$ for disky SFGs in S+S pairs., Comment: Accepted for publication in ApJS, 18 pages, 13 figures, 1 table

Online review systems are important components in influencing customers' purchase decisions. To manipulate a product's reputation, many stores hire large numbers of people to produce fake reviews to mislead customers. Previous methods tackle this problem by detecting malicious individuals, ignoring the fact that the spam activities are often formed in groups, where individuals work collectively to write fake reviews. Fake reviewer group detection, however, is more challenging due to the difficulties in capturing the underlying relationships in groups. In this work, we present an unsupervised and end-to-end approach for fake reviewer group detection in online reviews. Specifically, our method can be summarized into two procedures. First, cohensive groups are detected with modularity-based graph convolutional networks. Then the suspiciousness of each group is measured by several anomaly indicators from both individual and group levels. The fake reviewer groups can be finally detected through suspiciousness. Extensive experiments are conducted on real-world datasets, and the results show that our proposed method is effective in detecting fake reviewer groups compared with the state-of-the-art baselines.