Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userNeuroTreeNet: A New Method to Explore Horizontal Expansion Network
148
0
0.0
(
0
)
Added by
Shenlong Lou
Publication date
2018
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
It is widely recognized that the deeper networks or networks with more feature maps have better performance. Existing studies mainly focus on extending the network depth and increasing the feature maps of networks. At the same time, horizontal expansion network (e.g. Inception Model) as an alternative way to improve network performance has not been fully investigated. Accordingly, we proposed NeuroTreeNet (NTN), as a new horizontal extension network through the combination of random forest and Inception Model. Based on the tree structure, in which each branch represents a network and the root node features are shared to child nodes, network parameters are effectively reduced. By combining all features of leaf nodes, even less feature maps achieved better performance. In addition, the relationship between tree structure and the performance of NTN was investigated in depth. Comparing to other networks (e.g. VDSR_5) with equal magnitude parameters, our model showed preferable performance in super resolution reconstruction task.
rate research
Read More
We suggest to explore an entirely new method to experimentally and theoretically study the phase diagram of strongly interacting matter based on the triple nuclear collisions (TNC). We simulated the TNC using the UrQMD 3.4 model at the beam center-of-mass collision energies $sqrt{s_{NN}} = 200$ GeV and $sqrt{s_{NN}} = 2.76$ TeV. It is found that in the most central and simultaneous TNC the initial baryonic charge density is about 3 times higher than the one achieved in the usual binary nuclear collisions at the same energies. As a consequence, the production of protons and $Lambda$-hyperons is increased by a factor of 2 and 1.5, respectively. Using the MIT Bag model equation we study the evolution of the central cell in TNC and demonstrate that for the top RHIC energy of collision the baryonic chemical potential is 2-2.5 times larger than the one achieved in the binary nuclear collision at the same type of reaction. Based on these estimates, we show that TNC offers an entirely new possibility to study the QCD phase diagram at very high baryonic charge densities.
A patient-centric approach to healthcare leads to an informal social network among medical professionals. This chapter presents a research framework to: identify the collaboration structure among physicians that is effective and efficient for patients, discover effective structural attributes of a collaboration network that evolves during the course of providing care, and explore the impact of socio-demographic characteristics of healthcare professionals, patients, and hospitals on collaboration structures, from the point of view of measurable outcomes such as cost and quality of care. The framework uses illustrative examples drawn from a data set of patients undergoing hip replacement surgery.
Long-lived particles (LLPs) are a common feature in many beyond the Standard Model theories, including supersymmetry, and are generically produced in exotic Higgs decays. Unfortunately, no existing or proposed search strategy will be able to observe the decay of non-hadronic electrically neutral LLPs with masses above $sim$ GeV and lifetimes near the limit set by Big Bang Nucleosynthesis (BBN), $c tau lesssim 10^7 - 10^8$~m. We propose the MATHUSLA surface detector concept (MAssive Timing Hodoscope for Ultra Stable neutraL pArticles), which can be implemented with existing technology and in time for the high luminosity LHC upgrade to find such ultra-long-lived particles (ULLPs), whether produced in exotic Higgs decays or more general production modes. We also advocate for a dedicated LLP detector at a future 100 TeV collider, where a modestly sized underground design can discover ULLPs with lifetimes at the BBN limit produced in sub-percent level exotic Higgs decays.
Alongside the prevalence of mobile videos, the general public leans towards consuming vertical videos on hand-held devices. To revitalize the exposure of horizontal contents, we hereby set forth the exploration of automated horizontal-to-vertical (abbreviated as H2V) video conversion with our proposed H2V framework, accompanied by an accurately annotated H2V-142K dataset. Concretely, H2V framework integrates video shot boundary detection, subject selection and multi-object tracking to facilitate the subject-preserving conversion, wherein the key is subject selection. To achieve so, we propose a Rank-SS module that detects human objects, then selects the subject-to-preserve via exploiting location, appearance, and salient cues. Afterward, the framework automatically crops the video around the subject to produce vertical contents from horizontal sources. To build and evaluate our H2V framework, H2V-142K dataset is densely annotated with subject bounding boxes for 125 videos with 132K frames and 9,500 video covers, upon which we demonstrate superior subject selection performance comparing to traditional salient approaches, and exhibit promising horizontal-to-vertical conversion performance overall. By publicizing this dataset as well as our approach, we wish to pave the way for more valuable endeavors on the horizontal-to-vertical video conversion task.
This work presents a modular and hierarchical approach to learn policies for exploring 3D environments, called `Active Neural SLAM. Our approach leverages the strengths of both classical and learning-based methods, by using analytical path planners with learned SLAM module, and global and local policies. The use of learning provides flexibility with respect to input modalities (in the SLAM module), leverages structural regularities of the world (in global policies), and provides robustness to errors in state estimation (in local policies). Such use of learning within each module retains its benefits, while at the same time, hierarchical decomposition and modular training allow us to sidestep the high sample complexities associated with training end-to-end policies. Our experiments in visually and physically realistic simulated 3D environments demonstrate the effectiveness of our approach over past learning and geometry-based approaches. The proposed model can also be easily transferred to the PointGoal task and was the winning entry of the CVPR 2019 Habitat PointGoal Navigation Challenge.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
