GammaGAN: Gamma-Scaled Class Embeddings for Conditional Video Generation

In this paper, we propose a new model for conditional video generation (GammaGAN). Generally, it is challenging to generate a plausible video from a single image with a class label as a condition. Traditional methods based on conditional generative adversarial networks (cGANs) often encounter difficulties in effectively utilizing a class label, typically by concatenating a class label to the input or hidden layer. In contrast, the proposed GammaGAN adopts the projection method to effectively utilize a class label and proposes scaling class embeddings and normalizing outputs. Concretely, our proposed architecture consists of two streams: a class embedding stream and a data stream. In the class embedding stream, class embeddings are scaled to effectively emphasize class-specific differences. Meanwhile, the outputs in the data stream are normalized. Our normalization technique balances the outputs of both streams, ensuring a balance between the importance of feature vectors and class embeddings during training. This results in enhanced video quality. We evaluated the proposed method using the MUG facial expression dataset, which consists of six facial expressions. Compared with the prior conditional video generation model, ImaGINator, our model yielded relative improvements of 1.61%, 1.66%, and 0.36% in terms of PSNR, SSIM, and LPIPS, respectively. These results suggest potential for further advancements in conditional video generation.

Image Recommendation System Based on Environmental and Human Face Information

With the advancement of computer hardware and communication technologies, deep learning technology has made significant progress, enabling the development of systems that can accurately estimate human emotions. Factors such as facial expressions, gender, age, and the environment influence human emotions, making it crucial to understand and capture these intricate factors. Our system aims to recommend personalized images by accurately estimating human emotions, age, and gender in real time. The primary objective of our system is to enhance user experiences by recommending images that align with their current emotional state and characteristics. To achieve this, our system collects environmental information, including weather conditions and user-specific environment data through APIs and smartphone sensors. Additionally, we employ deep learning algorithms for real-time classification of eight types of facial expressions, age, and gender. By combining this facial information with the environmental data, we categorize the user's current situation into positive, neutral, and negative stages. Based on this categorization, our system recommends natural landscape images that are colorized using Generative Adversarial Networks (GANs). These recommendations are personalized to match the user's current emotional state and preferences, providing a more engaging and tailored experience. Through rigorous testing and user evaluations, we assessed the effectiveness and user-friendliness of our system. Users expressed satisfaction with the system's ability to generate appropriate images based on the surrounding environment, emotional state, and demographic factors such as age and gender. The visual output of our system significantly impacted users' emotional responses, resulting in a positive mood change for most users. Moreover, the system's scalability was positively received, with users acknowledging its potential benefits when installed outdoors and expressing a willingness to continue using it. Compared to other recommender systems, our integration of age, gender, and weather information provides personalized recommendations, contextual relevance, increased engagement, and a deeper understanding of user preferences, thereby enhancing the overall user experience. The system's ability to comprehend and capture intricate factors that influence human emotions holds promise in various domains, including human-computer interaction, psychology, and social sciences.

Efficient Multi-Scale Stereo-Matching Network Using Adaptive Cost Volume Filtering

While recent deep learning-based stereo-matching networks have shown outstanding advances, there are still some unsolved challenges. First, most state-of-the-art stereo models employ 3D convolutions for 4D cost volume aggregation, which limit the deployment of networks for resource-limited mobile environments owing to heavy consumption of computation and memory. Although there are some efficient networks, most of them still require a heavy computational cost to incorporate them to mobile computing devices in real-time. Second, most stereo networks indirectly supervise cost volumes through disparity regression loss by using the softargmax function. This causes problems in ambiguous regions, such as the boundaries of objects, because there are many possibilities for unreasonable cost distributions which result in overfitting problem. A few works deal with this problem by generating artificial cost distribution using only the ground truth disparity value that is insufficient to fully regularize the cost volume. To address these problems, we first propose an efficient multi-scale sequential feature fusion network (MSFFNet). Specifically, we connect multi-scale SFF modules in parallel with a cross-scale fusion function to generate a set of cost volumes with different scales. These cost volumes are then effectively combined using the proposed interlaced concatenation method. Second, we propose an adaptive cost-volume-filtering (ACVF) loss function that directly supervises our estimated cost volume. The proposed ACVF loss directly adds constraints to the cost volume using the probability distribution generated from the ground truth disparity map and that estimated from the teacher network which achieves higher accuracy. Results of several experiments using representative datasets for stereo matching show that our proposed method is more efficient than previous methods. Our network architecture consumes fewer parameters and generates reasonable disparity maps with faster speed compared with the existing state-of-the art stereo models. Concretely, our network achieves 1.01 EPE with runtime of 42 ms, 2.92M parameters, and 97.96G FLOPs on the Scene Flow test set. Compared with PSMNet, our method is 89% faster and 7% more accurate with 45% fewer parameters.

Semantic Segmentation Using Pixel-Wise Adaptive Label Smoothing via Self-Knowledge Distillation for Limited Labeling Data

To achieve high performance, most deep convolutional neural networks (DCNNs) require a significant amount of training data with ground truth labels. However, creating ground-truth labels for semantic segmentation requires more time, human effort, and cost compared with other tasks such as classification and object detection, because the ground-truth label of every pixel in an image is required. Hence, it is practically demanding to train DCNNs using a limited amount of training data for semantic segmentation. Generally, training DCNNs using a limited amount of data is problematic as it easily results in a decrease in the accuracy of the networks because of overfitting to the training data. Here, we propose a new regularization method called pixel-wise adaptive label smoothing (PALS) via self-knowledge distillation to stably train semantic segmentation networks in a practical situation, in which only a limited amount of training data is available. To mitigate the problem caused by limited training data, our method fully utilizes the internal statistics of pixels within an input image. Consequently, the proposed method generates a pixel-wise aggregated probability distribution using a similarity matrix that encodes the affinities between all pairs of pixels. To further increase the accuracy, we add one-hot encoded distributions with ground-truth labels to these aggregated distributions, and obtain our final soft labels. We demonstrate the effectiveness of our method for the Cityscapes dataset and the Pascal VOC2012 dataset using limited amounts of training data, such as 10%, 30%, 50%, and 100%. Based on various quantitative and qualitative comparisons, our method demonstrates more accurate results compared with previous methods. Specifically, for the Cityscapes test set, our method achieved mIoU improvements of 0.076%, 1.848%, 1.137%, and 1.063% for 10%, 30%, 50%, and 100% training data, respectively, compared with the method of the cross-entropy loss using one-hot encoding with ground truth labels.

Knowledge Distillation for Semantic Segmentation Using Channel and Spatial Correlations and Adaptive Cross Entropy

In this paper, we propose an efficient knowledge distillation method to train light networks using heavy networks for semantic segmentation. Most semantic segmentation networks that exhibit good accuracy are based on computationally expensive networks. These networks are not suitable for mobile applications using vision sensors, because computational resources are limited in these environments. In this view, knowledge distillation, which transfers knowledge from heavy networks acting as teachers to light networks as students, is suitable methodology. Although previous knowledge distillation approaches have been proven to improve the performance of student networks, most methods have some limitations. First, they tend to use only the spatial correlation of feature maps and ignore the relational information of their channels. Second, they can transfer false knowledge when the results of the teacher networks are not perfect. To address these two problems, we propose two loss functions: a channel and spatial correlation (CSC) loss function and an adaptive cross entropy (ACE) loss function. The former computes the full relationship of both the channel and spatial information in the feature map, and the latter adaptively exploits one-hot encodings using the ground truth labels and the probability maps predicted by the teacher network. To evaluate our method, we conduct experiments on scene parsing datasets: Cityscapes and Camvid. Our method presents significantly better performance than previous methods.

Depressive symptoms and risk of absence among workers in a manufacturing company: a 12-month follow-up study

Depression is a leading cause of reduced work ability and absence due to sickness. The objective of this study was to investigate how depressive symptoms are prospectively associated with subsequent absence, whether caused by illness or accidents, among manufacturing workers. This prospective study was conducted on 2,349 male and female employees that underwent a regular health examination at a university hospital. Depressive symptoms were measured at baseline using the Center for Epidemiologic Studies Depression (CES-D) Scale. Data on self-reported absence due to illness and accidents were obtained during a follow up of 1 yr. The incidences of sickness absence were 6.0% for men and 17.3% for women. Men and women with depressive symptoms (CES-D ≥16) were found to have higher odds of sickness absence during follow up (men: OR=4.06; 95% CI: 2.32-7.11; women: OR=1.75; 95% CI: 1.02-2.98), after adjustment for demographic and occupational factors. When depressive symptoms were divided into quartiles, significantly higher ORs of sickness absence were observed only among employees with the highest quartile of depressive symptoms. The study shows that depressive symptoms are a risk factor for future absence due to illness or accidents among manufacturing workers.

A Novel Cascade Classifier for Automatic Microcalcification Detection

In this paper, we present a novel cascaded classification framework for automatic detection of individual and clusters of microcalcifications (μC). Our framework comprises three classification stages: i) a random forest (RF) classifier for simple features capturing the second order local structure of individual μCs, where non-μC pixels in the target mammogram are efficiently eliminated; ii) a more complex discriminative restricted Boltzmann machine (DRBM) classifier for μC candidates determined in the RF stage, which automatically learns the detailed morphology of μC appearances for improved discriminative power; and iii) a detector to detect clusters of μCs from the individual μC detection results, using two different criteria. From the two-stage RF-DRBM classifier, we are able to distinguish μCs using explicitly computed features, as well as learn implicit features that are able to further discriminate between confusing cases. Experimental evaluation is conducted on the original Mammographic Image Analysis Society (MIAS) and mini-MIAS databases, as well as our own Seoul National University Bundang Hospital digital mammographic database. It is shown that the proposed method outperforms comparable methods in terms of receiver operating characteristic (ROC) and precision-recall curves for detection of individual μCs and free-response receiver operating characteristic (FROC) curve for detection of clustered μCs.

Forest Walk Methods for Localizing Body Joints from Single Depth Image

We present multiple random forest methods for human pose estimation from single depth images that can operate in very high frame rate. We introduce four algorithms: random forest walk, greedy forest walk, random forest jumps, and greedy forest jumps. The proposed approaches can accurately infer the 3D positions of body joints without additional information such as temporal prior. A regression forest is trained to estimate the probability distribution to the direction or offset toward the particular joint, relative to the adjacent position. During pose estimation, the new position is chosen from a set of representative directions or offsets. The distribution for next position is found from traversing the regression tree from new position. The continual position sampling through 3D space will eventually produce an expectation of sample positions, which we estimate as the joint position. The experiments show that the accuracy is higher than current state-of-the-art pose estimation methods with additional advantage in computation time.

Structure of the catalytic phosphatase domain of MTMR8: implications for dimerization, membrane association and reversible oxidation

Myotubularin-related proteins are a large family of phosphoinositide phosphatases; their activity, stability and subcellular localization are regulated by dimeric interactions with other members of the family. Here, the crystal structure of the phosphatase domain of MTMR8 is reported. Conformational deviation of the two loops that mediate interaction with the PH-GRAM domain suggests that the PH-GRAM domain interacts differently with the phosphatase domain of each MTMR member. The protein exists as a dimer with twofold symmetry, providing insight into a novel mode of dimerization mediated by the phosphatase domain. Structural comparison and mutation studies suggest that Lys255 of MTMR8 interacts with the substrate diacylglycerol moiety, similar to Lys333 of MTMR2, although the positions of these residues are different. The catalytic activity of the MTMR8 phosphatase domain is inhibited by oxidation and is reversibly reactivated by reduction, suggesting the presence of an oxidation-protective intermediate other than a disulfide bond owing to the absence of a cysteine within a disulfide-bond distance from Cys338.

Licochalcone A, a polyphenol present in licorice, suppresses UV-induced COX-2 expression by targeting PI3K, MEK1, and B-Raf

Licorice is a traditional botanical medicine, and has historically been commonly prescribed in Asia to treat various diseases. Glycyrrhizin (Gc), a triterpene compound, is the most abundant phytochemical constituent of licorice. However, high intake or long-term consumption of Gc has been associated with a number of side effects, including hypertension. However, the presence of alternative bioactive compounds in licorice with anti-carcinogenic effects has long been suspected. Licochalcone A (LicoA) is a prominent member of the chalcone family and can be isolated from licorice root. To date, there have been no reported studies on the suppressive effect of LicoA against solar ultraviolet (sUV)-induced cyclooxygenase (COX)-2 expression and the potential molecular mechanisms involved. Here, we show that LicoA, a major chalcone compound of licorice, effectively inhibits sUV-induced COX-2 expression and prostaglandin E2 PGE₂ generation through the inhibition of activator protein 1 AP-1 transcriptional activity, with an effect that is notably more potent than Gc. Western blotting analysis shows that LicoA suppresses sUV-induced phosphorylation of Akt/ mammalian target of rapamycin (mTOR) and extracellular signal-regulated kinases (ERK)1/2/p90 ribosomal protein S6 kinase (RSK) in HaCaT cells. Moreover, LicoA directly suppresses the activity of phosphoinositide 3-kinase (PI3K), mitogen-activated protein kinase kinase (MEK)1, and B-Raf, but not Raf-1 in cell-free assays, indicating that PI3K, MEK1, and B-Raf are direct molecular targets of LicoA. We also found that LicoA binds to PI3K and B-Raf in an ATP-competitive manner, although LicoA does not appear to compete with ATP for binding with MEK1. Collectively, these results provide insight into the biological action of LicoA, which may have potential for development as a skin cancer chemopreventive agent.

Joint depth map and color consistency estimation for stereo images with different illuminations and cameras

Abstract—In this paper, we propose a method that infers both accurate depth maps and color-consistent stereo images for radiometrically varying stereo images. In general, stereo matching and performing color consistency between stereo images are a chicken-and-egg problem since it is not a trivial task to simultaneously achieve both goals. Hence, we have developed an iterative framework in which these two processes can boost each other. First, we transform the input color images to log-chromaticity color space, from which a linear relationship can be established during constructing a joint pdf of transformed left and right color images. From this joint pdf, we can estimate a linear function that relates the corresponding pixels in stereo images. Based on this linear property, we present a new stereo matching cost by combining Mutual Information (MI), SIFT descriptor, and segment-based plane-fitting to robustly find correspondence for stereo image pairs which undergo radiometric variations. Meanwhile, we devise a Stereo Color Histogram Equalization (SCHE) method to produce color-consistent stereo image pairs, which conversely boost the disparity map estimation. Experimental results show that our method produces both accurate depth maps and color-consistent stereo images, even for stereo images with severe radiometric differences.

Robust stereo matching using adaptive normalized cross-correlation

A majority of the existing stereo matching algorithms assume that the corresponding color values are similar to each other. However, it is not so in practice as image color values are often affected by various radiometric factors such as illumination direction, illuminant color, and imaging device changes. For this reason, the raw color recorded by a camera should not be relied on completely, and the assumption of color consistency does not hold good between stereo images in real scenes. Therefore, the performance of most conventional stereo matching algorithms can be severely degraded under the radiometric variations. In this paper, we present a new stereo matching measure that is insensitive to radiometric variations between left and right images. Unlike most stereo matching measures, we use the color formation model explicitly in our framework and propose a new measure, called the Adaptive Normalized Cross-Correlation (ANCC), for a robust and accurate correspondence measure. The advantage of our method is that it is robust to lighting geometry, illuminant color, and camera parameter changes between left and right images, and does not suffer from the fattening effect unlike conventional Normalized Cross-Correlation (NCC). Experimental results show that our method outperforms other state-of-the-art stereo methods under severely different radiometric conditions between stereo images.

Cytotoxic flavonoids as agonists of peroxisome proliferator-activated receptor gamma on human cervical and prostate cancer cells

We conducted in silico screening for human peroxisome proliferator-activated receptor gamma (hPPARgamma) by performing an automated docking study with 450 flavonoids. Among the eight flavonoids as possible agonists of hPPARgamma, only 3,6-dihydroxyflavone (4) increased the binding between PPARgamma and steroid receptor coactivator-1 (SRC-1), approximately 5-fold, and showed one order higher binding affinity for PPARgamma than a reference compound, indomethacin. The 6-hydroxy group of the A-ring of 3,6-dihydroxyflavone (4) participated in hydrogen-bonding interactions with the side chain of Tyr327, His449, and Tyr473. The B-ring formed a hydrophobic interaction with Leu330, Leu333, Val339, Ile341, and Met364. Therefore, 3,6-dihydroxyflavone is a potent agonist of hPPAR with cytotoxic effects on human cervical and prostate cancer cells.

Dynamic microfunnel culture enhances mouse embryo development and pregnancy rates

BACKGROUND

Despite advances in in vitro manipulation of preimplantation embryos, there is still a reduction in the quality of embryos produced leading to lower pregnancy rates compared with embryos produced in vivo. We hypothesized that a dynamic microfunnel embryo culture system would enhance outcomes by better mimicking the fluid-mechanical and biochemical stimulation embryos experience in vivo from ciliary currents and oviductal contractions.

METHODS AND RESULTS

Mouse embryos were cultured in microdrop-static control, microfunnel-static control or microfunnel-dynamic conditions with microfluidics. All groups tested had greater than 90% total blastocyst development from zygotes after 96 h culture. Blastocyst developmental stage was significantly enhanced (P < 0.01) under dynamic microfunnel culture conditions as evidenced by an increased percentage of hatching or hatched blastocysts (Microdrop-control 31%; Microfunnel-control 23%; Microfunnel-pulsatile 71%) and significantly higher (P < 0.01) average number of cells per blastocyst (Microdrop-control 67 +/- 3; Microfunnel-control 60 +/- 3; Microfunnel-pulsatile 109 +/- 5). Blastocyst cell numbers in dynamic microfunnel cultures (109 +/- 5) more closely matched numbers obtained from in vivo grown blastocysts (144 +/- 9). Importantly, dynamic microfunnel culture significantly improved embryo implantation and ongoing pregnancy rates over static culture to levels approaching that of in utero derived preimplantation embryos.

CONCLUSIONS

The improved pregnancy outcomes along with the simple and user-friendly design of the microfluidic/microfunnel system has potential to alleviate many inefficiencies in embryo production for biomedical research, genetic gain in domestic species and assisted reproductive technologies in humans.

The resveratrol analogue 3,5,3',4',5'-pentahydroxy-trans-stilbene inhibits cell transformation via MEK

Resveratrol, present in grapes and red wine, is reported to be a natural chemopreventive agent against cancer. However, the concentrations required to exert these effects may be difficult to achieve by drinking only 1 or 2 glasses of red wine a day. Therefore, developing more potent, nontoxic analogues of resveratrol may provide a feasible means of achieving an effective physiologic concentration. Here we report that the resveratrol analogue, 3,5,3',4',5'-pentahydroxy-trans-stilbene (RSVL2), inhibits 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced neoplastic transformation in JB6 P+ mouse epidermal cells. Further, we identified MEK/ERK signaling as the direct molecular target for the anticancer effects of RSVL2 and demonstrated that RSVL2 inhibited MEK1, but not Raf1 or ERK2 kinase activity. RSVL2 also dose-dependently suppressed MEK1 kinase activity induced by TPA and the inhibition of H-Ras-induced cell transformation was much stronger for RSVL2 than for PD098059 or resveratrol. Both in vitro and ex vivo pull-down assays indicated that RSVL2, but not resveratrol, directly bound with GST-MEK1, but did not compete with ATP for binding. Docking data indicated that the low inhibitory activity of resveratrol might be due to the lack of the hydroxyl group at the meta position of the B ring, thereby preventing resveratrol from forming a hydrogen bond with the backbone amide group of Ser212, which is the key interaction for stabilizing the inactive conformation of the activation loop.

Expression, purification, crystallization, and preliminary X-Ray analysis of the human UDP-glucose dehydrogenase

UDP-glucose dehydrogenase (UGDH) catalyzes the synthesis of UDP-glucuronic acid from UDP-glucose resulting in the formation of proteoglycans that are involved in promoting normal cellular growth and migration. Overproduction of proteoglycans has been implicated in the progression of certain epithelial cancers. Here, human UGDH (hUGDH) was purified and crystallized from a solution of 0.2 M ammonium sulfate, 0.1 M Na cacodylate, pH 6.5, and 21% PEG 8000. Diffraction data were collected to a resolution of 2.8 A. The crystal belongs to the orthorhombic space group P2(1)2(1)2(1) with unit-cell parameters a = 173.25, b = 191.16, c = 225.94 A, and alpha = beta = gamma = 90.0 degrees. Based on preliminary analysis of the diffraction data, we propose that the biological unit of hUGDH is a tetramer.

A study on plasma prostaglandin E2 levels in hepatitis B carriers and patients with chronic active hepatitis

Prostaglandin E₂ (PGE₂), one of the major prostaglandins synthesized in human monocyte and macrophage, is able to modulate T lymphocyte reactivity, such as lymphokine secretion and cytotoxicity. Some immunologic abnormalities such as alteration in the synthesis of PGE₂ by monocyte and macrophage or in the response of T lymphocytes to PGE₂ can be found in clinical disease. We measured the plasma PGE₂ level in the control group and patients with chronic liver disease. The results were obtained as follows. 1. The mean plasma PGE₂ level was 2.65 ± 0.69 pg/ml in the control group. 2. The mean plasma PGE₂ level was 9.07 ± 5.89 pg/ml in 15 patients with chronic active hepatitis and was significantly higher than that of the control group (p<0.01). 3. The plasma mean PGE₂ level was 4.65 ± 1.59 pg/ml in 8 patients in the healing stage or stable stage of chronic hepatitis and was tend to decrease. However, this decrease is significantly different from that of the control group. 4. The plasma PGE₂ level was 3.5 ± 0.92 pg/ml in 4 hepatitis B carriers and was not significantly different from that of the control group (p<0.05). This results suggest that plasma PGE₂ can be used for the measurement of cell-mediated immunity and follow-up study in patients with chronic active hepatitis and hepatitis B carriers.