Search

Parametric 3D models have enabled a wide variety of computer vision and graphics tasks, such as modeling human faces, bodies and hands. In 3D face modeling, 3DMM is the most widely used parametric model, but can't generate fine geometric details solely from identity and expression inputs. To tackle this limitation, we propose a neural parametric model named DNPM for the facial geometric details, which utilizes deep neural network to extract latent codes from facial displacement maps encoding details and wrinkles. Built upon DNPM, a novel 3DMM named Detailed3DMM is proposed, which augments traditional 3DMMs by including the synthesis of facial details only from the identity and expression inputs. Moreover, we show that DNPM and Detailed3DMM can facilitate two downstream applications: speech-driven detailed 3D facial animation and 3D face reconstruction from a degraded image. Extensive experiments have shown the usefulness of DNPM and Detailed3DMM, and the progressiveness of two proposed applications.

Speech-driven facial animation methods usually contain two main classes, 3D and 2D talking face, both of which attract considerable research attention in recent years. However, to the best of our knowledge, the research on 3D talking face does not go deeper as 2D talking face, in the aspect of lip-synchronization (lip-sync) and speech perception. To mind the gap between the two sub-fields, we propose a learning framework named Learn2Talk, which can construct a better 3D talking face network by exploiting two expertise points from the field of 2D talking face. Firstly, inspired by the audio-video sync network, a 3D sync-lip expert model is devised for the pursuit of lip-sync between audio and 3D facial motion. Secondly, a teacher model selected from 2D talking face methods is used to guide the training of the audio-to-3D motions regression network to yield more 3D vertex accuracy. Extensive experiments show the advantages of the proposed framework in terms of lip-sync, vertex accuracy and speech perception, compared with state-of-the-arts. Finally, we show two applications of the proposed framework: audio-visual speech recognition and speech-driven 3D Gaussian Splatting based avatar animation.

Text input is a desired feature for AR glasses. While there already exist various input modalities (e.g., voice, mid-air gesture), the diverse demands required by different input scenarios can hardly be met by the small number of fixed input postures offered by existing solutions. In this paper, we present Handwriting Velcro, a novel text input solution for AR glasses based on flexible touch sensors. The distinct advantage of our system is that it can easily stick to different body parts, thus endowing AR glasses with posture-adaptive handwriting input. We explored the design space of on-body device positions and identified the best interaction positions for various user postures. To flatten users' learning curves, we adapt our device to the established writing habits of different users by training a 36-character (i.e., A-Z, 0-9) recognition neural network in a human-in-the-loop manner. Such a personalization attempt ultimately achieves a low error rate of 0.005 on average for users with different writing styles. Subjective feedback shows that our solution has a good performance in system practicability and social acceptance. Empirically, we conducted a heuristic study to explore and identify the best interaction Position-Posture Correlation. Experimental results show that our Handwriting Velcro excels similar work [6] and commercial product in both practicality (12.3 WPM) and user-friendliness in different contexts. © 2023 ACM.

The last few decades have witnessed an emerging trend of wearable soft sensors; however, there are important signal-processing challenges for soft sensors that still limit their practical deployment. They are error-prone when displaced, resulting in significant deviations from their ideal sensor output. In this work, we propose a novel prototype that integrates an elbow pad with a sparse network of soft sensors. Our prototype is fully bio-compatible, stretchable, and wearable. We develop a learning-based method to predict the elbow orientation angle and achieve an average tracking error of 9.82 degrees for single-user multi-motion experiments. With transfer learning, our method achieves the average tracking errors of 10.98 degrees and 11.81 degrees across different motion types and users, respectively. Our core contributions lie in a solution that realizes robust and stable human joint motion tracking across different device displacements. © 2023 ACM.

The last few decades have witnessed an emerging trend of wearable soft sensors; however, there are important signal-processing challenges for soft sensors that still limit their practical deployment. They are error-prone when displaced, resulting in significant deviations from their ideal sensor output. In this work, we propose a novel prototype that integrates an elbow pad with a sparse network of soft sensors. Our prototype is fully bio-compatible, stretchable, and wearable. We develop a learning-based method to predict the elbow orientation angle and achieve an average tracking error of 9.82 degrees for single-user multi-motion experiments. With transfer learning, our method achieves the average tracking errors of 10.98 degrees and 11.81 degrees across different motion types and users, respectively. Our core contributions lie in a solution that realizes robust and stable human joint motion tracking across different device displacements., Comment: 25 pages, 14 figures

Interior scene colorization is vastly demanded in areas such as personalized architecture design. Existing works either require manual efforts to colorize individual objects or conform to fixed color patterns automatically learned from prior knowledge, whilst neglecting user preference. Quantitatively identifying user preferences is challenging, particularly at the early stage of the design process. The 3D setup also presents new challenges as the inhabitant can observe from any possible viewpoint. We propose a representative view selection method based on visual attention and a progressive preference inference model. We particularly focus on the progressive integration of eye-tracked user preference, which enables the assistance in creativity support and allows the possibility of convergent thinking. A series of user studies have been conducted to validate the effectiveness of the proposed view selection method, preference inference model and the creativity support mechanism.

Deep neural networks are being increasingly used for short-term traffic flow prediction, which can be generally categorized as convolutional (CNNs) or graph neural networks (GNNs). CNNs are preferable for region-wise traffic prediction by taking advantage of localized spatial correlations, whilst GNNs achieves better performance for graph-structured traffic data. When applied to region-wise traffic prediction, CNNs typically partition an underlying territory into grid-like spatial units, and employ standard convolutions to learn spatial dependence among the units. However, standard convolutions with fixed geometric structures cannot fully model the nonstationary characteristics of local traffic flows. To overcome the deficiency, we introduce deformable convolution that augments the spatial sampling locations with additional offsets, to enhance the modeling capability of spatial nonstationarity. On this basis, we design a deep deformable convolutional residual network, namely DeFlow-Net, that can effectively model global spatial dependence, local spatial nonstationarity, and temporal periodicity of traffic flows. Furthermore, to better fit with convolutions, we suggest to first aggregate traffic flows according to pre-conceived regions or self-organized regions based on traffic flows, then dispose to sequentially organized raster images for network input. Extensive experiments on real-world traffic flows demonstrate that DeFlow-Net outperforms GNNs and existing CNNs using standard convolutions, and spatial partition by pre-conceived regions or self-organized regions further enhances the performance. We also demonstrate the advantage of DeFlow-Net in maintaining spatial autocorrelation, and reveal the impacts of partition shapes and scales on deep traffic flow prediction.

Deep learning methods are being increasingly used for urban traffic prediction where spatiotemporal traffic data is aggregated into sequentially organized matrices that are then fed into convolution-based residual neural networks. However, the widely known modifiable areal unit problem within such aggregation processes can lead to perturbations in the network inputs. This issue can significantly destabilize the feature embeddings and the predictions, rendering deep networks much less useful for the experts. This paper approaches this challenge by leveraging unit visualization techniques that enable the investigation of many-to-many relationships between dynamically varied multi-scalar aggregations of urban traffic data and neural network predictions. Through regular exchanges with a domain expert, we design and develop a visual analytics solution that integrates 1) a Bivariate Map equipped with an advanced bivariate colormap to simultaneously depict input traffic and prediction errors across space, 2) a Morans I Scatterplot that provides local indicators of spatial association analysis, and 3) a Multi-scale Attribution View that arranges non-linear dot plots in a tree layout to promote model analysis and comparison across scales. We evaluate our approach through a series of case studies involving a real-world dataset of Shenzhen taxi trips, and through interviews with domain experts. We observe that geographical scale variations have important impact on prediction performances, and interactive visual exploration of dynamically varying inputs and outputs benefit experts in the development of deep traffic prediction models.