We propose a novel 3D-aware diffusion-based method for generating photorealistic talking head videos directly from a single identity image and explicit control signals (e.g., expressions). Our method generates Multiplane Images (MPIs) that ensure geometric consistency, making them ideal for immersive viewing experiences like binocular videos for VR headsets. Unlike existing methods that often require a separate stage or joint optimization to reconstruct a 3D representation (such as NeRF or 3D Gaussians), our approach directly generates the final output through a single denoising process, eliminating the need for post-processing steps to render novel views efficiently. To effectively learn from monocular videos, we introduce a training mechanism that reconstructs the output MPI randomly in either the target or the reference camera space. This approach enables the model to simultaneously learn sharp image details and underlying 3D information. Extensive experiments demonstrate the effectiveness of our method, which achieves competitive avatar quality and novel-view rendering capabilities, even without explicit 3D reconstruction or high-quality multi-view training data.

Video generation models have demonstrated great capability of producing impressive monocular videos, however, the generation of 3D stereoscopic video remains under-explored. We propose a pose-free and training-free approach for generating 3D stereoscopic videos using an off-the-shelf monocular video generation model. Our method warps a generated monocular video into camera views on stereoscopic baseline using estimated video depth, and employs a novel frame matrix video inpainting framework. The framework leverages the video generation model to inpaint frames observed from different timestamps and views. This effective approach generates consistent and semantically coherent stereoscopic videos without scene optimization or model fine-tuning. Moreover, we develop a disocclusion boundary re-injection scheme that further improves the quality of video inpainting by alleviating the negative effects propagated from disoccluded areas in the latent space. We validate the efficacy of our proposed method by conducting experiments on videos from various generative models, including Sora [4], Lumiere [2], WALT [8], and Zeroscope [12]. The experiments demonstrate that our method has a significant improvement over previous methods. Project page at https://daipengwa.github.io/SVG_ProjectPage/

Quick Response (QR) codes have become increasingly popular for screen-camera communication due to their swift readability and widespread smartphone use. Nevertheless, they are vulnerable to privacy invasions from unauthorized photography. Addressing this, we propose a novel Moiré encryption technique-based secure screen-camera communication system, named MoiréComm. The Moiré encryption can enhance security by using distinct spatial frequency patterns for camouflage. The original QR code is revealed as a Moiré pattern only when the camera in a designated position, e.g., directly in front and 30 cm from the screen. From any other positions, only the camouflaged QR code can be seen. Decryption schemes are customized for different scenarios. The multi-frame approach achieves a decryption success of over 98.6% within 13.2 frames in handheld scenarios. Conditional generative adversarial network (cGAN)-based decryption method decodes the Moiré QR code images with a 98.8% success rate in 0.02 s within three frames and is also applicable in handheld scenarios. For fixed screen-camera setups, our fast decryption scheme achieves 99.4% success within two frames, with average 0.4 s latency. Significantly, the decryption rate plunges to 0% for surveillance cameras displaced by 20$^\circ$or more than$\ge$10 cm from the target position, demonstrating MoiréComm's resilience against attacks.

Virtual environments (VEs) are pivotal for virtual, augmented, and mixed reality systems. Despite advances in 3D generation and reconstruction, the direct creation of 3D objects within an established 3D scene (represented as NeRF) for novel VE creation remains a relatively unexplored domain. This process is complex, requiring not only the generation of high-quality 3D objects but also their seamless integration into the existing scene. To this end, we propose a novel pipeline featuring an intuitive interface, dubbed GO-NeRF. Our approach takes text prompts and user-specified regions as inputs and leverages the scene context to generate 3D objects within the scene. We employ a compositional rendering formulation that effectively integrates the generated 3D objects into the scene, utilizing optimized 3D-aware opacity maps to avoid unintended modifications to the original scene. Furthermore, we develop tailored optimization objectives and training strategies to enhance the model's ability to capture scene context and mitigate artifacts, such as floaters, that may occur while optimizing 3D objects within the scene. Extensive experiments conducted on both forward-facing and 360°scenes demonstrate the superior performance of our proposed method in generating objects that harmonize with surrounding scenes and synthesizing high-quality novel view images. The code will be at https://daipengwa.github.io/G0-NeRF/.

Remote video conferencing systems (RVCS) are widely adopted in personal and professional communication. However, they often lack the co-presence experience of in-person meetings. This is largely due to the absence of intuitive visual cues and clear spatial relationships among remote participants, which can lead to speech interruptions and loss of attention. This paper presents ChatDirector, a novel RVCS that overcomes these limitations by incorporating space-aware visual presence and speech-aware attention transition assistance. ChatDirector employs a real-time pipeline that converts participants’ RGB video streams into 3D portrait avatars and renders them in a virtual 3D scene. We also contribute a decision tree algorithm that directs the avatar layouts and behaviors based on participants’ speech states. We report on results from a user study (N=16) where we evaluated ChatDirector. The satisfactory algorithm performance and complimentary subject user feedback imply that ChatDirector significantly enhances communication efficacy and user engagement.

3D head avatars built with neural implicit volumetric representations have achieved unprecedented levels of pho-torealism. However, the computational cost of these methods remains a significant barrier to their widespread adoption, particularly in real-time applications such as virtual reality and teleconferencing. While attempts have been made to develop fast neural rendering approaches for static scenes, these methods cannot be simply employed to support realistic facial expressions, such as in the case of a dynamic facial performance. To address these challenges, we propose a novel fast 3D neural implicit head avatar model that achieves real-time rendering while maintaining fine-grained controllability and high rendering quality. Our key idea lies in the introduction of local hash table blendshapes, which are learned and attached to the vertices of an underlying face parametric model. These per-vertex hash-tables are linearly merged with weights predicted via a CNN, re-sulting in expression dependent embeddings. Our novel representation enables efficient density and color predictions using a lightweight MLP, which is further accelerated by a hierarchical nearest neighbor search method. Extensive experiments show that our approach runs in real-time while achieving comparable rendering quality to state-of-the-arts and decent results on challenging expressions.

The demand for ultra-high-resolution imaging in mobile phone photography is continuously increasing. However, the image resolution of mobile devices is typically constrained by the size of the CMOS sensor. Although deep learning-based super-resolution (SR) techniques have the potential to overcome this limitation, existing SR neural network models require large computational resources, making them unsuitable for real-time SR imaging on current mobile devices. Additionally, cloud-based SR systems pose privacy leakage risks. In this paper, we propose M3Cam, an innovative and lightweight SR imaging system for mobile phones. M3Cam can ensure high-quality 16× SR image (4× in both height and width) visualization with almost negligible latency. In detail, we utilize an optical image stabilization (OIS) module for lens control and introduce a new modality of data, namely gyroscope readings, to achieve high-precision and compact optical flow estimation modules. Building upon this concept, we design a multi-frame-based SR model utilizing the Swin Transformer. Our proposed system can generate a 16× SR image from four captured low-resolution images in real-time, with low computational load, low inference latency, and minimal reliance on runtime RAM. Through extensive experiments, we demonstrate that our proposed multi-modal optical flow model significantly enhances pixel alignment accuracy between multiple frames and delivers outstanding 16× SR imaging results under various shooting scenarios. Code and dataset are available at: https://github.com/liangjindeamo-yuer/M3CAM

We propose a method to learn a high-quality implicit 3D head avatar from a monocular RGB video captured in the wild. The learnt avatar is driven by a parametric face model to achieve user-controlled facial expressions and head poses. Our hybrid pipeline combines the geometry prior and dynamic tracking of a 3DMM with a neural radiance field to achieve fine-grained control and photorealism. To reduce over-smoothing and improve out-of-model expressions synthesis, we propose to predict local features anchored on the 3DMM geometry. These learnt features are driven by 3DMM deformation and interpolated in 3D space to yield the volumetric radiance at a designated query point. We further show that using a Convolutional Neural Network in the UV space is critical in incorporating spatial context and producing representative local features. Extensive experiments show that we are able to reconstruct high-quality avatars, with more accurate expression-dependent details, good generalization to out-of-training expressions, and quantitatively superior renderings compared to other state-of-the-art approaches.

Multi-frame super-resolution methods can generate high resolution images by combining multiple captures of the same scene; however, the performance of merged results are susceptible to degradation due to a lack of precision in image registration. In this study, we sought to develop a robust multi-frame super resolution method (called OISSR) for use on smartphone cameras with a optical image stabilizer (OIS). Acoustic injection is used to alter the readings from the built-in MEMS gyroscope to control the lens motion in the OIS module (note that the image sensor is fixed). We employ a priori knowledge of the induced lens motion to facilitate optical flow estimation with sub-pixel accuracy, and the output high-precision pixel alignment vectors are utilized to merge the multiple frames to reconstruct the final super resolution image. Extensive experiments on a OISSR prototype implemented on a Xiaomi 10Ultra demonstrate the high performance and effectiveness of the proposed system in obtaining the quadruple enhanced resolution imaging.

Optical image stabilizers (OIS) are widely used in digital cameras to counteract motion blur caused by camera shakes in capturing videos and photos. In this paper, we sought to expand the applicability of the lens-shift OIS technology for metric depth estimation, i.e., let a RGB camera to achieve the similar function of a time-of-flight (ToF) camera. Instead of having to move the entire camera for depth estimation, we propose DoCam, which controls the lens motion in the OIS module to achieve 3D reconstruction. After controlling the lens motion by altering the MEMS gyroscopes readings through acoustic injection, we improve the traditional bundle adjustment algorithm by establishing additional constraints from the linearity of the lens control model for high-precision camera pose estimation. Then, we elaborate a dense depth reconstruction algorithm to compute depth maps at real-world scale from multiple captures with micro lens motion (i.e., ≤ 3 mm). Extensive experiments demonstrate that our proposed DoCam can enable a 2D color camera to estimate high-accuracy depth information of the captured scene by means of controlling lens motion in the OIS. DoCam is suitable for a variety of applications that require depth information of the scenes, especially when only a single color camera is available and located at a fixed position.

This study reveals that on-board hardware modules leak electromagnetic (EM) emissions whenever audio or camera data is accessed, and proposes Magdefender scheme that explores the possibility of using the magnetometer built into mobile devices to detect eavesdropping instances by malicious apps and even the unscrupulous phone vendors. However, the target EM signals generated by accessing multimedia data is weak and tends to be buried beneath other noisy EM signals from apps running in the foreground. It is also subject to the external interference from geomagnetic signals generated by the device movement. To cope with the challenges, we adopt a generative adversarial networks (GAN) based model to facilitate the extraction of target EM signals indicating the occurrence of eavesdropping from the overall magnetometer readings. We also develop a neural network-based classifier with triplet loss embedding to identify the EM signals from the camera and/or microphones. Empirical results demonstrate the efficacy of MagDefenderin recognizing instances of eavesdropping on cameras/microphones data, with average accuracy of 97.3% when applied to the trained devices, and average 91.5% on unseen mobile devices.

In this paper, we address the problem of building dense correspondences between human images under arbitrary camera viewpoints and body poses. Prior art either assumes small motion between frames or relies on local descriptors, which cannot handle large motion or visually ambiguous body parts, e.g., left vs. right hand. In contrast, we propose a deep learning framework that maps each pixel to a feature space, where the feature distances reflect the geodesic distances among pixels as if they were projected onto the surface of a 3D human scan. To this end, we introduce novel loss functions to push features apart according to their geodesic distances on the surface. Without any semantic annotation, the proposed embeddings automatically learn to differentiate visually similar parts and align different subjects into an unified feature space. Extensive experiments show that the learned embeddings can produce accurate correspondences between images with remarkable generalization capabilities on both intra and inter subjects.1

The deep multi-view stereo (MVS) and stereo matching approaches generally construct 3D cost volumes to regularize and regress the output depth or disparity. These methods are limited when high-resolution outputs are needed since the memory and time costs grow cubically as the volume resolution increases. In this paper, we propose a both memory and time efficient cost volume formulation that is complementary to existing multi-view stereo and stereo matching approaches based on 3D cost volumes. First, the proposed cost volume is built upon a standard feature pyramid encoding geometry and context at gradually finer scales. Then, we can narrow the depth (or disparity) range of each stage by the depth (or disparity) map from the previous stage. With gradually higher cost volume resolution and adaptive adjustment of depth (or disparity) intervals, the output is recovered in a coarser to fine manner. We apply the cascade cost volume to the representative MVS-Net, and obtain a 35.6% improvement on DTU benchmark (1st place), with 50.6% and 59.3% reduction in GPU memory and run-time. It is also the state-of-the-art learning-based method on Tanks and Temples benchmark. The statistics of accuracy, run-time and GPU memory on other representative stereo CNNs also validate the effectiveness of our proposed method. Our source code is available at https://github.com/alibaba/cascade-stereo.

Previous methods on estimating detailed human depth often require supervised training with ‘ground truth’ depth data. This paper presents a self-supervised method that can be trained on YouTube videos without known depth, which makes training data collection simple and improves the generalization of the learned network. The self-supervised learning is achieved by minimizing a photo-consistency loss, which is evaluated between a video frame and its neighboring frames warped according to the estimated depth and the 3D non-rigid motion of the human body. To solve this non-rigid motion, we first estimate a rough SMPL model at each video frame and compute the non-rigid body motion accordingly, which enables self-supervised learning on estimating the shape details. Experiments demonstrate that our method enjoys better generalization, and performs much better on data in the wild.

This paper presents a neural network to estimate a detailed depth map of the foreground human in a single RGB image. The result captures geometry details such as cloth wrinkles, which are important in visualization applications. To achieve this goal, we separate the depth map into a smooth base shape and a residual detail shape and design a network with two branches to regress them respectively. We design a training strategy to ensure both base and detail shapes can be faithfully learned by the corresponding network branches. Furthermore, we introduce a novel network layer to fuse a rough depth map and surface normals to further improve the final result. Quantitative comparison with fused `ground truth' captured by real depth cameras and qualitative examples on unconstrained Internet images demonstrate the strength of the proposed method.

This paper presents a novel polarimetric dense monocular SLAM (PDMS) algorithm based on a polarization camera. The algorithm exploits both photometric and polarimetric light information to produce more accurate and complete geometry. The polarimetric information allows us to recover the azimuth angle of surface normals from each video frame to facilitate dense reconstruction, especially at textureless or specular regions. There are two challenges in our approach: 1) surface azimuth angles from the polarization camera are very noisy; and 2) we need a near real-time solution for SLAM. Previous successful methods on polarimetric multi-view stereo are offline and require manually pre-segmented object masks to suppress the effects of erroneous angle information along boundaries. Our fully automatic approach efficiently iterates azimuth-based depth propagations, two-view depth consistency check, and depth optimization to produce a depthmap in real-time, where all the algorithmic steps are carefully designed to enable a GPU implementation. To our knowledge, this paper is the first to propose a photometric and polarimetric method for dense SLAM. We have qualitatively and quantitatively evaluated our algorithm against a few of competing methods, demonstrating the superior performance on various indoor and outdoor scenes.