Remote photoplethysmography (rPPG) monitors heart rate without requiring physical contact, which allows for a wide variety of applications. Deep learning-based rPPG have demonstrated superior performance over the traditional approaches in controlled
context. However, the lighting situation in indoor space is typically complex, with uneven light distribution and frequent variations in illumination. It lacks a fair comparison of different methods under different illuminations using the same dataset. In this paper, we present a public dataset, namely the BH-rPPG dataset, which contains data from twelve subjects under three illuminations: low, medium, and high illumination. We also provide the ground truth heart rate measured by an oximeter. We evaluate the performance of three deep learning-based methods to that of four traditional methods using two public datasets: the UBFC-rPPG dataset and the BH-rPPG dataset. The experimental results demonstrate that traditional methods are generally more resistant to fluctuating illuminations. We found that the rPPGNet achieves lowest MAE among deep learning-based method under medium illumination, whereas the CHROM achieves 1.5 beats per minute (BPM), outperforming the rPPGNet by 60%. These findings suggest that while developing deep learning-based heart rate estimation algorithms, illumination variation should be taken into account. This work serves as a benchmark for rPPG performance evaluation and it opens a pathway for future investigation into deep learning-based rPPG under illumination variations.
Constructing and animating humans is an important component for building virtual worlds in a wide variety of applications such as virtual reality or robotics testing in simulation. As there are exponentially many variations of humans with different s
hape, pose and clothing, it is critical to develop methods that can automatically reconstruct and animate humans at scale from real world data. Towards this goal, we represent the pedestrians shape, pose and skinning weights as neural implicit functions that are directly learned from data. This representation enables us to handle a wide variety of different pedestrian shapes and poses without explicitly fitting a human parametric body model, allowing us to handle a wider range of human geometries and topologies. We demonstrate the effectiveness of our approach on various datasets and show that our reconstructions outperform existing state-of-the-art methods. Furthermore, our re-animation experiments show that we can generate 3D human animations at scale from a single RGB image (and/or an optional LiDAR sweep) as input.
Sensor simulation is a key component for testing the performance of self-driving vehicles and for data augmentation to better train perception systems. Typical approaches rely on artists to create both 3D assets and their animations to generate a new
scenario. This, however, does not scale. In contrast, we propose to recover the shape and motion of pedestrians from sensor readings captured in the wild by a self-driving car driving around. Towards this goal, we formulate the problem as energy minimization in a deep structured model that exploits human shape priors, reprojection consistency with 2D poses extracted from images, and a ray-caster that encourages the reconstructed mesh to agree with the LiDAR readings. Importantly, we do not require any ground-truth 3D scans or 3D pose annotations. We then incorporate the reconstructed pedestrian assets bank in a realistic LiDAR simulation system by performing motion retargeting, and show that the simulated LiDAR data can be used to significantly reduce the amount of annotated real-world data required for visual perception tasks.
Few-shot object detection is a challenging but realistic scenario, where only a few annotated training images are available for training detectors. A popular approach to handle this problem is transfer learning, i.e., fine-tuning a detector pretraine
d on a source-domain benchmark. However, such transferred detector often fails to recognize new objects in the target domain, due to low data diversity of training samples. To tackle this problem, we propose a novel Context-Transformer within a concise deep transfer framework. Specifically, Context-Transformer can effectively leverage source-domain object knowledge as guidance, and automatically exploit contexts from only a few training images in the target domain. Subsequently, it can adaptively integrate these relational clues to enhance the discriminative power of detector, in order to reduce object confusion in few-shot scenarios. Moreover, Context-Transformer is flexibly embedded in the popular SSD-style detectors, which makes it a plug-and-play module for end-to-end few-shot learning. Finally, we evaluate Context-Transformer on the challenging settings of few-shot detection and incremental few-shot detection. The experimental results show that, our framework outperforms the recent state-of-the-art approaches.
Person re-identification (re-id), the process of matching pedestrian images across different camera views, is an important task in visual surveillance. Substantial development of re-id has recently been observed, and the majority of existing models a
re largely dependent on color appearance and assume that pedestrians do not change their clothes across camera views. This limitation, however, can be an issue for re-id when tracking a person at different places and at different time if that person (e.g., a criminal suspect) changes his/her clothes, causing most existing methods to fail, since they are heavily relying on color appearance and thus they are inclined to match a person to another person wearing similar clothes. In this work, we call the person re-id under clothing change the cross-clothes person re-id. In particular, we consider the case when a person only changes his clothes moderately as a first attempt at solving this problem based on visible light images; that is we assume that a person wears clothes of a similar thickness, and thus the shape of a person would not change significantly when the weather does not change substantially within a short period of time. We perform cross-clothes person re-id based on a contour sketch of person image to take advantage of the shape of the human body instead of color information for extracting features that are robust to moderate clothing change. Due to the lack of a large-scale dataset for cross-clothes person re-id, we contribute a new dataset that consists of 33698 images from 221 identities. Our experiments illustrate the challenges of cross-clothes person re-id and demonstrate the effectiveness of our proposed method.
Generative models, especially Generative Adversarial Networks (GANs), have received significant attention recently. However, it has been observed that in terms of some attributes, e.g. the number of simple geometric primitives in an image, GANs are n
ot able to learn the target distribution in practice. Motivated by this observation, we discover two specific problems of GANs leading to anomalous generalization behaviour, which we refer to as the sample insufficiency and the pixel-wise combination. For the first problem of sample insufficiency, we show theoretically and empirically that the batchsize of the training samples in practice may be insufficient for the discriminator to learn an accurate discrimination function. It could result in unstable training dynamics for the generator, leading to anomalous generalization. For the second problem of pixel-wise combination, we find that besides recognizing the positive training samples as real, under certain circumstances, the discriminator could be fooled to recognize the pixel-wise combinations (e.g. pixel-wise average) of the positive training samples as real. However, those combinations could be visually different from the real samples in the target distribution. With the fooled discriminator as reference, the generator would obtain biased supervision further, leading to the anomalous generalization behaviour. Additionally, in this paper, we propose methods to mitigate the anomalous generalization of GANs. Extensive experiments on benchmark show our proposed methods improve the FID score up to 30% on natural image dataset.
Exploration of structure-property relationships as a function of dopant concentration is commonly based on mean field theories for solid solutions. However, such theories that work well for semiconductors tend to fail in materials with strong correla
tions, either in electronic behavior or chemical segregation. In these cases, the details of atomic arrangements are generally not explored and analyzed. The knowledge of the generative physics and chemistry of the material can obviate this problem, since defect configuration libraries as stochastic representation of atomic level structures can be generated, or parameters of mesoscopic thermodynamic models can be derived. To obtain such information for improved predictions, we use data from atomically resolved microscopic images that visualize complex structural correlations within the system and translate them into statistical mechanical models of structure formation. Given the significant uncertainties about the microscopic aspects of the materials processing history along with the limited number of available images, we combine model optimization techniques with the principles of statistical hypothesis testing. We demonstrate the approach on data from a series of atomically-resolved scanning transmission electron microscopy images of Mo$_x$Re$_{1-x}$S$_2$ at varying ratios of Mo/Re stoichiometries, for which we propose an effective interaction model that is then used to generate atomic configurations and make testable predictions at a range of concentrations and formation temperatures.
The exceptional electronic, optical and chemical properties of two-dimensional materials strongly depend on the 3D atomic structure and crystal defects. Using Re-doped MoS2 as a model, here we develop scanning atomic electron tomography (sAET) to det
ermine the 3D atomic positions and crystal defects such as dopants, vacancies and ripples with a precision down to 4 picometers. We measure the 3D bond distortion and local strain tensor induced by single dopants for the first time. By directly providing experimental 3D atomic coordinates to density functional theory (DFT), we obtain more truthful electronic band structures than those derived from conventional DFT calculations relying on relaxed 3D atomic models, which is confirmed by photoluminescence measurements. We anticipate that sAET is not only generally applicable to the determination of the 3D atomic coordinates of 2D materials, heterostructures and thin films, but also could transform ab initio calculations by using experimental 3D atomic coordinates as direct input to better predict and discover new physical, chemical and electronic properties.
Recent years have witnessed great success of convolutional neural network (CNN) for various problems both in low and high level visions. Especially noteworthy is the residual network which was originally proposed to handle high-level vision problems
and enjoys several merits. This paper aims to extend the merits of residual network, such as skip connection induced fast training, for a typical low-level vision problem, i.e., single image super-resolution. In general, the two main challenges of existing deep CNN for supper-resolution lie in the gradient exploding/vanishing problem and large numbers of parameters or computational cost as CNN goes deeper. Correspondingly, the skip connections or identity mapping shortcuts are utilized to avoid gradient exploding/vanishing problem. In addition, the skip connections have naturally centered the activation which led to better performance. To tackle with the second problem, a lightweight CNN architecture which has carefully designed width, depth and skip connections was proposed. In particular, a strategy of gradually varying the shape of network has been proposed for residual network. Different residual architectures for image super-resolution have also been compared. Experimental results have demonstrated that the proposed CNN model can not only achieve state-of-the-art PSNR and SSIM results for single image super-resolution but also produce visually pleasant results. This paper has extended the mmm 2017 oral conference paper with a considerable new analyses and more experiments especially from the perspective of centering activations and ensemble behaviors of residual network.
