Neural 3D Scene Reconstruction with Indoor Planar Priors.

This paper addresses the challenge of reconstructing 3D indoor scenes from multi-view images. Many previous works have shown impressive reconstruction results on textured objects, but they still have difficulty in handling low-textured planar regions, which are common in indoor scenes. An approach to solving this issue is to incorporate planar constraints into the depth map estimation in multi-view stereo-based methods, but the per-view plane estimation and depth optimization lack both efficiency and multi-view consistency. In this work, we show that the planar constraints can be conveniently integrated into the recent implicit neural representation-based reconstruction methods. Specifically, we use an MLP network to represent the signed distance function as the scene geometry. Based on the Manhattan-world assumption and the Atlanta-world assumption, planar constraints are employed to regularize the geometry in floor and wall regions predicted by a 2D semantic segmentation network. To resolve the inaccurate segmentation, we encode the semantics of 3D points with another MLP and design a novel loss that jointly optimizes the scene geometry and semantics in 3D space. Experiments on ScanNet and 7-Scenes datasets show that the proposed method outperforms previous methods by a large margin on 3D reconstruction quality. The code and supplementary materials are available at https://zju3dv.github.io/ manhattan sdf.

Animatable Implicit Neural Representations for Creating Realistic Avatars From Videos.

This paper addresses the challenge of reconstructing an animatable human model from a multi-view video. Some recent works have proposed to decompose a non-rigidly deforming scene into a canonical neural radiance field and a set of deformation fields that map observation-space points to the canonical space, thereby enabling them to learn the dynamic scene from images. However, they represent the deformation field as translational vector field or SE(3) field, which makes the optimization highly under-constrained. Moreover, these representations cannot be explicitly controlled by input motions. Instead, we introduce blend weight fields to produce the deformation fields. Based on the skeleton-driven deformation, blend weight fields are used with 3D human skeletons to generate observation-to-canonical and canonical-to-observation correspondences. Since 3D human skeletons are more observable, they can regularize the learning of deformation fields. Moreover, the blend weight fields can be combined with input skeletal motions to generate new deformation fields to animate the human model. To improve the quality of human modeling, we further represent the human geometry as a signed distance field in the canonical space. Additionally, a neural point displacement field is introduced to enhance the capability of the blend weight field on modeling detailed human motions. Experiments show that our approach significantly outperforms recent human modeling methods.

Implicit Neural Representations With Structured Latent Codes for Human Body Modeling.

This paper addresses the challenge of novel view synthesis for a human performer from a very sparse set of camera views. Some recent works have shown that learning implicit neural representations of 3D scenes achieves remarkable view synthesis quality given dense input views. However, the representation learning will be ill-posed if the views are highly sparse. To solve this ill-posed problem, our key idea is to integrate observations over video frames. To this end, we propose Neural Body, a new human body representation which assumes that the learned neural representations at different frames share the same set of latent codes anchored to a deformable mesh, so that the observations across frames can be naturally integrated. The deformable mesh also provides geometric guidance for the network to learn 3D representations more efficiently. Furthermore, we combine Neural Body with implicit surface models to improve the learned geometry. To evaluate our approach, we perform experiments on both synthetic and real-world data, which show that our approach outperforms prior works by a large margin on novel view synthesis and 3D reconstruction. We also demonstrate the capability of our approach to reconstruct a moving person from a monocular video on the People-Snapshot dataset.

Magnetically driven micro-optical choppers fabricated by two-photon polymerization.

In this Letter, a series of magnetically driven micro-optical choppers based on customized photoresist were fabricated by two-photon polymerization (TPP) technology. Synthetic Fe3O4 nanoparticles (NPs) were modified and dispersed in the original photoresist to achieve magnetic field response. After accurately formulating a magnetic photoresist containing Rhodamine B to reduce the light transmittance, four micro-optical choppers with different slot widths were printed using optimized processing parameters. The micro-optical choppers were remotely manipulated to rotate by the external magnetic field. More importantly, the function demonstration of the micro-optical choppers with an excellent chopping effect was achieved at a given light wavelength of 515 nm. The magnetically driven micro-optical choppers provide a new approach, to the best of our knowledge, for the fabrication of external field-responsive optical components.

NerfCap: Human Performance Capture With Dynamic Neural Radiance Fields.

This paper addresses the challenge of human performance capture from sparse multi-view or monocular videos. Given a template mesh of the performer, previous methods capture the human motion by non-rigidly registering the template mesh to images with 2D silhouettes or dense photometric alignment. However, the detailed surface deformation cannot be recovered from the silhouettes, while the photometric alignment suffers from instability caused by appearance variation in the videos. To solve these problems, we propose NerfCap, a novel performance capture method based on the dynamic neural radiance field (NeRF) representation of the performer. Specifically, a canonical NeRF is initialized from the template geometry and registered to the video frames by optimizing the deformation field and the appearance model of the canonical NeRF. To capture both large body motion and detailed surface deformation, NerfCap combines linear blend skinning with embedded graph deformation. In contrast to the mesh-based methods that suffer from fixed topology and texture, NerfCap is able to flexibly capture complex geometry and appearance variation across the videos, and synthesize more photo-realistic images. In addition, NerfCap can be pre-trained end to end in a self-supervised manner by matching the synthesized videos with the input videos. Experimental results on various datasets show that NerfCap outperforms prior works in terms of both surface reconstruction accuracy and novel-view synthesis quality.

MPS-NeRF: Generalizable 3D Human Rendering From Multiview Images.

There has been rapid progress recently on 3D human rendering, including novel view synthesis and pose animation, based on the advances of neural radiance fields (NeRF). However, most existing methods focus on person-specific training and their training typically requires multi-view videos. This paper deals with a new challenging task - rendering novel views and novel poses for a person unseen in training, using only multiview still images as input without videos. For this task, we propose a simple yet surprisingly effective method to train a generalizable NeRF with multiview images as conditional input. The key ingredient is a dedicated representation combining a canonical NeRF and a volume deformation scheme. Using a canonical space enables our method to learn shared properties of human and easily generalize to different people. Volume deformation is used to connect the canonical space with input and target images and query image features for radiance and density prediction. We leverage the parametric 3D human model fitted on the input images to derive the deformation, which works quite well in practice when combined with our canonical NeRF. The experiments on both real and synthetic data with the novel view synthesis and pose animation tasks collectively demonstrate the efficacy of our method.

PVNet: Pixel-Wise Voting Network for 6DoF Object Pose Estimation.

This paper addresses the problem of instance-level 6DoF object pose estimation from a single RGB image. Many recent works have shown that a two-stage approach, which first detects keypoints and then solves a Perspective-n-Point (PnP) problem for pose estimation, achieves remarkable performance. However, most of these methods only localize a set of sparse keypoints by regressing their image coordinates or heatmaps, which are sensitive to occlusion and truncation. Instead, we introduce a Pixel-wise Voting Network (PVNet) to regress pixel-wise vectors pointing to the keypoints and use these vectors to vote for keypoint locations. This creates a flexible representation for localizing occluded or truncated keypoints. Another important feature of this representation is that it provides uncertainties of keypoint locations that can be further leveraged by the PnP solver. Experiments show that the proposed approach outperforms the state of the art on the LINEMOD, Occluded LINEMOD, YCB-Video, and Tless datasets, while being efficient for real-time pose estimation. We further create a Truncated LINEMOD dataset to validate the robustness of our approach against truncation. The code is available at https://github.com/zju3dv/pvnet.

Circular RNA circ-CCT3 promotes bortezomib resistance in multiple myeloma via modulating miR-223-3p/BRD4 axis.

Multiple myeloma is a frequent hematologic malignancy. Bortezomib is the first-line drug for multiple myeloma chemotherapy. The present study aimed to investigate the potential role and mechanism of circular RNA chaperonin containing TCP1 subunit 3 (circ-CCT3) in bortezomib resistance of multiple myeloma. The levels of circ-CCT3, microRNA-223-3p (miR-223-3p), and bromodomain-containing 4 (BRD4) were detected by quantitative real-time PCR or western blot. Cell Counting Kit-8 (CCK-8) method was used to measure the half-inhibitory concentration of bortezomib and cell viability. Cell cycle distribution, apoptosis, proliferation and migration were determined by flow cytometry, 5-ethynyl-2'-deoxyuridine, and wound healing assay. The levels of relevant proteins were checked via western blot. The binding association between miR-223-3p and circ-CCT3/BRD4 was validated via a dual-luciferase reporter assay. Circ-CCT3 and BRD4 were upregulated, while miR-223-3p was downregulated in bortezomib-resistant multiple myeloma patients and cells. Silencing of circ-CCT3 enhanced the sensitivity of bortezomib-resistant multiple myeloma cells to bortezomib. Circ-CCT3 knockdown weakened bortezomib resistance via modulating miR-223-3p. Moreover, miR-223-3p increased bortezomib sensitivity by inhibiting BRD4. Downregulation of circ-CCT3 attenuated bortezomib resistance of multiple myeloma via regulating miR-223-3p/BRD4 pathway, which provided a new potential target for multiple myeloma chemoresistance.

CTSB Knockdown Inhibits Proliferation and Tumorigenesis in HL-60 Cells.

Background: Cathepsin B (CTSB) was well documented in solid tumors, up-regulated of CTSB expression is linked with progression of tumors. However, the study of CTSB in adult leukemia has not been reported. Methods: Total RNA was isolated from PBMC (peripheral blood mononuclear cell) of AML patients and healthy donors. qRT-PCR was performed to detect the expression of CTSB. The association of CTSB expression with the patients' overall survival (OS) and disease-free survival (DFS) were analyzed. Stable HL-60 CTSB-shRNA cell lines were established by retrovirus infection and puromycin selection. Cell proliferation was detected by CCK-8 analysis. Tumorigenesis ability was analyzed by soft agar and xenograft nude mice model. Western blot was performed to detect the expression of CTSB and the proteins of cell signaling pathway. Results: The mRNA expression level of CTSB was up-regulated in AML patients compared to healthy control (p<0.001), and CTSB expression was significantly higher in M1, M2, M4 and M5 AML samples than healthy control. The CTSB expression in AML was associated with WBC count (p=0.037). Patients with high CTSB expression had a relatively poor OS (p=0.007) and a shorter DFS (p=0.018). Moreover, the expression level of CTSB may act as an independent prognostic factor for both OS (p=0.011) and DFS (p=0.004). Knockdown CTSB expression in HL-60 cells could inhibit the cells' proliferation and tumorigeneses in vitro and in vivo. Further study showed knockdown CTSB expression in HL-60 cells could inactive the AKT signaling pathway. Conclusions: CTSB mRNA was upregulated in AML patients. CTSB overexpression was correlated with poor prognosis and may serve as an independent prognostic factor for both OS and DFS in AML patients. Knockdown CTSB expression in HL-60 cells could inhibit the cells' proliferation and tumorigenesis. The underlying mechanism may be the inhibition of the AKT signaling pathway.

Mostly harmless regulation? Electronic cigarettes, public policy, and consumer welfare.

Electronic cigarettes are a less harmful alternative to combustible cigarettes. We analyze data on e-cigarette choices in an online experimental market. Our data and mixed logit model capture two sources of consumer optimization errors: overestimates of the relative risks of e-cigarettes and present bias. Our novel data and policy analysis make three contributions. First, our predictions about e-cigarette use under counterfactual policy scenarios provide new information about current regulatory tradeoffs. Second, we provide empirical evidence about the role consumer optimization errors play in tobacco product choices. Third, we contribute to behavioral welfare analysis of policies that address individual optimization errors. Compared with standard cost-benefit analysis, our behavioral welfare economics analysis leads to much larger estimates of the costs of policies that discourage e-cigarette use or the benefits of policies that encourage e-cigarette use.

INHBA knockdown inhibits proliferation and invasion of nasopharyngeal carcinoma SUNE1 cells in vitro.

Up-regulated expression of INHBA has been reported in multiple malignant tumors. However, in nasopharyngeal carcinoma (NPC), the expression pattern and clinical significance of INHBA are still unclear. This study aimed to detect the expression of INHBA and its prognostic significance in NPC, then explore the tumor-associated functions of INHBA gene and the potential mechanism. The INHBA expression of mRNA levels in tumor tissues and noncancerous nasopharyngeal tissues was investigated by RT-qPCR. The protein expression in cells were detected by western blot. Cell proliferation was detected by CCK assay and cell invasion ability was evaluated by Transwell assay. The expression of INHBA in paraffin-embedded NPC tissues was detected by immunohistochemistry (IHC). Statistical analyses were further applied to assess the clinical significance of INHBA expression. The result reveals INHBA mRNA level is elevated in NPC tissues compared to those in noncancerous nasopharyngeal epithelial tissues. In paraffin-embedded NPC tissues, immunoreactivity of INHBA was primarily detected in 53.70% (58/108) of these patients. The overexpression was notably associated with the clinical stage (UICC) (P=0.048), N classification (P=0.042), carotid sheath involvement (P=0.016), and decreased disease-free survival (DFS) (P=0.004) and overall survival (OS) (P=0.010). Multivariate analysis revealed that INHBA expression was an independent prognostic factor for DFS (P=0.028). CCK assay showed SUNE1 cells' proliferation was decreased in INHBA knockdown group than control. Transwell assay showed the invasion of SUNE1 cells was decreased in INHBA knockdown group by comparison with control. Further study showed knockdown of INHBA expression in SUNE1 cells could block the TGF-ß signaling pathway. In conclusion, INHBA is up-regulated in NPC, and is significantly correlated with clinical stage (UICC), N stage, carotid sheath involvement, and survival. Knockdown INHBA in SUNE1 cells could inhibit the cells' proliferation and invasion. The underlying mechanism may be blockade of the TGF-ß signaling pathway.

Expression status of cyclase­associated protein 2 as a prognostic marker for human breast cancer.

Cyclase-associated protein 2 (CAP2) protein is reported to be upregulated in hepatocellular carcinoma (HCC). However, data regarding its expression pattern and clinical relevance in breast cancer are unknown. The aim of this study was to investigate CAP2 expression and its prognostic significance in breast cancer. CAP2 expression at the mRNA and protein levels was examined by real­time quantitative-polymerase chain reaction and western blotting in 10 paired breast cancer tissues and adjacent normal tissues. The expression level of CAP2 protein in normal breast epithelial cells and breast cancer cell lines was quantified by western blotting. CAP2 protein expression was analyzed in paraffin­embedded breast cancer samples, paired adjacent non­tumor and normal breast tissues by immunohistochemical analysis. Statistical analyses were also performed to evaluate the clinicopathological significance of CAP2 expression. The results showed that the expression of CAP2 mRNA and protein was higher in breast cancer than that noted in the adjacent normal tissues in 10 paired samples. The expression level of CAP2 protein in breast cancer cell lines was higher than that in normal breast epithelial cells. In paraffin­embedded tissue samples, the expression of CAP2 was higher in breast cancer than that found in the adjacent non­cancerous tissues and normal breast tissues. Compared with the adjacent non­cancerous tissues, overexpression of CAP2 was detected in 29.4% (37/126) of the patients. Overexpression of CAP2 was significantly associated with progesterone receptor (PR) expression (p<0.05), and decreased overall survival (OS) (p<0.05). In multivariate analysis, expression of CAP2 was an independent prognostic factor for OS [hazard ratio (HR), 4.821; 95% confidence interval (CI), 2.442­9.518; p<0.001]. CAP2 is upregulated in breast cancer and is associated with the expression of PR and patient survival. CAP2 may serve as a prognostic indicator for patients with breast cancer.