Existing methods on video-based action recognition are generally view-dependent, i.e., performing recognition from the same views seen in the training data. We present a novel multiview ...spatio-temporal and-or graph (MST-AOG) representation for cross-view action recognition, i.e., the recognition is performed on the video from an unknown and unseen view. As a compositional model, MST-AOG compactly represents the hierarchical combinatorial structures of cross-view actions by explicitly modeling the geometry, appearance and motion variations. This paper proposes effective methods to learn the structure and parameters of MST-AOG. The inference based on MST-AOG enables action recognition from novel views. The training of MST-AOG takes advantage of the 3D human skeleton data obtained from Kinect cameras to avoid annotating enormous multi-view video frames, which is error-prone and time-consuming, but the recognition does not need 3D information and is based on 2D video input. A new Multiview Action3D dataset has been created and will be released. Extensive experiments have demonstrated that this new action representation significantly improves the accuracy and robustness for cross-view action recognition on 2D videos.
We propose a systematic learning-based approach to the generation of massive quantities of synthetic 3D scenes and arbitrary numbers of photorealistic 2D images thereof, with associated ground truth ...information, for the purposes of training, benchmarking, and diagnosing learning-based computer vision and robotics algorithms. In particular, we devise a learning-based pipeline of algorithms capable of automatically generating and rendering a potentially infinite variety of indoor scenes by using a stochastic grammar, represented as an attributed Spatial And-Or Graph, in conjunction with state-of-the-art physics-based rendering. Our pipeline is capable of synthesizing scene layouts with high diversity, and it is configurable inasmuch as it enables the precise customization and control of important attributes of the generated scenes. It renders photorealistic RGB images of the generated scenes while automatically synthesizing detailed, per-pixel ground truth data, including visible surface depth and normal, object identity, and material information (detailed to object parts), as well as environments (e.g., illuminations and camera viewpoints). We demonstrate the value of our synthesized dataset, by improving performance in certain machine-learning-based scene understanding tasks—depth and surface normal prediction, semantic segmentation, reconstruction, etc.—and by providing benchmarks for and diagnostics of trained models by modifying object attributes and scene properties in a controllable manner.
Full text
Available for:
CEKLJ, DOBA, EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, IZUM, KILJ, KISLJ, MFDPS, NLZOH, NUK, OBVAL, OILJ, PILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, SIK, UILJ, UKNU, UL, UM, UPUK, VKSCE, ZAGLJ
We present a human-centric method to sample and synthesize 3D room layouts and 2D images thereof, to obtain large-scale 2D/3D image data with the perfect per-pixel ground truth. An attributed spatial ...And-Or graph (S-AOG) is proposed to represent indoor scenes. The S-AOG is a probabilistic grammar model, in which the terminal nodes are object entities including room, furniture, and supported objects. Human contexts as contextual relations are encoded by Markov Random Fields (MRF) on the terminal nodes. We learn the distributions from an indoor scene dataset and sample new layouts using Monte Carlo Markov Chain. Experiments demonstrate that the proposed method can robustly sample a large variety of realistic room layouts based on three criteria: (i) visual realism comparing to a state-of-the-art room arrangement method, (ii) accuracy of the affordance maps with respect to ground-truth, and (ii) the functionality and naturalness of synthesized rooms evaluated by human subjects.
Hepatocellular carcinoma (HCC) is an aggressive subtype of liver cancer with few effective treatments, and the underlying mechanisms that drive HCC pathogenesis remain poorly characterized. ...Identifying genes and pathways essential for HCC cell growth will aid the development of new targeted therapies for HCC. Using a kinome CRISPR screen in three human HCC cell lines, we identified transformation/transcription domain‐associated protein (TRRAP) as an essential gene for HCC cell proliferation. TRRAP has been implicated in oncogenic transformation, but how it functions in cancer cell proliferation is not established. Here, we show that depletion of TRRAP or its co‐factor, histone acetyltransferase KAT5, inhibits HCC cell growth through induction of p53‐independent and p21‐independent senescence. Integrated cancer genomics analyses using patient data and RNA sequencing identified mitotic genes as key TRRAP/KAT5 targets in HCC, and subsequent cell cycle analyses revealed that TRRAP‐depleted and KAT5‐depleted cells are arrested at the G2/M phase. Depletion of topoisomerase II alpha (TOP2A), a mitotic gene and TRRAP/KAT5 target, was sufficient to recapitulate the senescent phenotype of TRRAP/KAT5 knockdown. Conclusion: Our results uncover a role for TRRAP/KAT5 in promoting HCC cell proliferation by activating mitotic genes. Targeting the TRRAP/KAT5 complex is a potential therapeutic strategy for HCC.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SAZU, SBCE, SBMB, UL, UM, UPUK
With the development of reinforced concrete bridges, the fatigue problem of concrete has attracted extensive attention. Failure occurs when concrete is subjected to cyclic compression that is less ...than the static compressive strength of concrete after a certain amount of cyclic loading (fatigue life). Compared to traditional methods with shortage of low accuracy, resulting uncertain safety problem, to address this issue, we utilized gradient boosting regression tree (GBRT) and multiple linear regression (MLR) to predict the fatigue life of concrete under repeated compression based on the collated 275 sets of experimental data. All prediction results were compared with that calculated by the existing European codes. Results analysis reveal that both proposed GBRT and MLR models have lower errors and higher correlation coefficients than that of existing formula/code models. Furthermore, the prediction accuracy of the GBRT model can be improved by using the existing classical formulas to reconstruct the features and selecting suitable features by ranking the feature importance, without increasing the number of features. Experimental results reveal that GBRT model can achieve the best performance with lowest RMSE and highest
R
2
. Moreover, the sensitivity analysis revealed that the proposed GBRT model is insensitive to input characteristics of fatigue life, which can be used for accurate calculation and design analysis of concrete compressive fatigue life, providing important reference value for formulating concrete technical design specification based on fatigue life.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, ZAGLJ
Increasing evidence has shown that microRNAs (miRNAs) play a significant functional role by directly regulating respective targets in cancer stem cell (CSC)-induced non-small cell lung cancer (NSCLC) ...progression and resistance to therapy. In this study, we found that hsa-miR-124a was downregulated during spheroid formation of the NSCLC cell lines SPC-A1 and NCI-H1650 and NSCLC tissues compared with normal lung cells and tissues. Patients with lower hsa-miR-124a expression had shorter overall survival (OS) and progression free survival (PFS). Moreover, ubiquitin-specific protease 14 (USP14) was confirmed to be a direct target of hsa-miR-124a. Furthermore, concomitant low hsa-miR-124a expression and high USP14 expression were correlated with a shorter median OS and PFS in NSCLC patients. Cellular functional analysis verified that the tumor suppressor hsa-miR-124a negatively regulated cell growth and self-renewal, and promoted apoptosis and gefitinib sensitivity of lung cancer stem cells by suppressing its target gene USP14. Our results provide the first evidence that USP14 is a direct target of hsa-miR-124a, and that hsa-miR-124a inhibits stemness and enhances the gefitinib sensitivity of NSCLC cells by targeting USP14. Thus, hsa-miR-124a and USP14 may be useful as tumor biomarkers for the diagnosis and treatment of NSCLC.
•Hsa-miR-124a regulates growth, self-renewal and gefitinib sensitivity of LCSCs by suppressing USP14.•USP14 is a direct target of hsa-miR-124a.•NSCLC patients with high USP14 expression and low hsa-miR-124a expression had significantly decreased OS and PFS.•Hsa-miR-124a and USP14 may be useful as tumor biomarkers for the diagnosis and treatment of NSCLC.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK, ZRSKP
The ten-eleven translocation 1 (TET1) gene is the founding member of the TET family of enzymes (TET1/2/3) that convert 5-methylcytosine to 5-hydroxymethylcytosine. Although TET1 was first identified ...as a fusion partner of the mixed lineage leukemia (MLL) gene in acute myeloid leukemia carrying t(10,11), its definitive role in leukemia is unclear. In contrast to the frequent down-regulation (or loss-of-function mutations) and critical tumor-suppressor roles of the three TET genes observed in various types of cancers, here we show that TET1 is a direct target of MLL-fusion proteins and is significantly up-regulated in MLL -rearranged leukemia, leading to a global increase of 5-hydroxymethylcytosine level. Furthermore, our both in vitro and in vivo functional studies demonstrate that Tet1 plays an indispensable oncogenic role in the development of MLL -rearranged leukemia, through coordination with MLL-fusion proteins in regulating their critical cotargets, including homeobox A9 (Hoxa9)/myeloid ecotropic viral integration 1 (Meis1)/pre-B-cell leukemia homeobox 3 (Pbx3) genes. Collectively, our data delineate an MLL-fusion/Tet1/Hoxa9/Meis1/Pbx3 signaling axis in MLL -rearranged leukemia and highlight TET1 as a potential therapeutic target in treating this presently therapy-resistant disease.
Full text
Available for:
BFBNIB, NMLJ, NUK, PNG, SAZU, UL, UM, UPUK
Background and Aims
Hepatocellular carcinoma (HCC) is among the most common cancer types worldwide, yet patients with HCC have limited treatment options. There is an urgent need to identify drug ...targets that specifically inhibit the growth of HCC cells.
Approach and Results
We used a CRISPR library targeting ~2,000 druggable genes to perform a high‐throughput screen and identified adenylosuccinate lyase (ADSL), a key enzyme involved in the de novo purine synthesis pathway, as a potential drug target for HCC. ADSL has been implicated as a potential oncogenic driver in some cancers, but its role in liver cancer progression remains unknown. CRISPR‐mediated knockout of ADSL impaired colony formation of liver cancer cells by affecting AMP production. In the absence of ADSL, the growth of liver tumors is retarded in vivo. Mechanistically, we found that ADSL knockout caused S‐phase cell cycle arrest not by inducing DNA damage but by impairing mitochondrial function. Using data from patients with HCC, we also revealed that high ADSL expression occurs during tumorigenesis and is linked to poor survival rate.
Conclusions
Our findings uncover the role of ADSL‐mediated de novo purine synthesis in fueling mitochondrial ATP production to promote liver cancer cell growth. Targeting ADSL may be a therapeutic approach for patients with HCC.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SAZU, SBCE, SBMB, UL, UM, UPUK
The crop intercepted photosynthetically active radiation (PAR) and radiation use efficiency (RUE) vary markedly in different intercropping systems. The HHLA (horizontally homogeneous leaf area) and ...ERCRT (extended row crop radiation transmission) models have been established to calculate the intercepted PAR for intercrops. However, there is still a lack of study on the intercepted PAR and RUE under different intercropping configurations using different models. To evaluate the intercepted PAR and RUE in maize and soybean under different intercropping systems, we tested different strip intercropping configurations (SI1, SI2, and SI3 based on ERCRT model) and a row intercropping configurations (RI based on HHLA model) in comparison to monoculture. Our results showed that the intercepted PAR and RUE of intercropping systems were all higher than those of monoculture. The soybean intercepted PAR in strip intercropping was 1.35 times greater than that in row intercropping. In row intercropping (RI), the lack of soybean intercepted PAR resulted in a significant reduction of soybean dry matter. Therefore, it is not the recommended configuration for soybean. In strip intercropping patterns, with the distance between maize strip increased by 0.2 m, the intercepted PAR of soybean increased by 20%. The SI2 (maize row spacing at 0.4 m and the distance between maize strip at 1.6 m) was the recommended configuration to achieve the highest value of intercepted PAR and RUE among tested strip intercropping configurations. The method of dry matter estimation using intercepted PAR and RUE is useful in simulated experiments. The simulated value was verified in comparison with experimental data, which confirmed the credibility of the simulation model. Moreover, it also provides help in the development of functional-structural plant model (FSPM).
Full text
Available for:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
We propose a notion of affordance that takes into account physical quantities generated when the human body interacts with real-world objects, and introduce a learning framework that incorporates the ...concept of human utilities, which in our opinion provides a deeper and finer-grained account not only of object affordance but also of people's interaction with objects. Rather than defining affordance in terms of the geometric compatibility between body poses and 3D objects, we devise algorithms that employ physicsbased simulation to infer the relevant forces/pressures acting on body parts. By observing the choices people make in videos (particularly in selecting a chair in which to sit) our system learns the comfort intervals of the forces exerted on body parts (while sitting). We account for people's preferences in terms of human utilities, which transcend comfort intervals to account also for meaningful tasks within scenes and spatiotemporal constraints in motion planning, such as for the purposes of robot task planning.