31 skills found · Page 1 of 2
Robbings / Chatgpt Graph NavigatorTransform ChatGPT into a navigable knowledge graph. Visualize complex branches with the spatial Graph View, manage history via the Git-style Timeline Tree, and enjoy a growing toolkit of workflow utilities.
ZJUFanLab / SpaTalkKnowledge-graph-based cell-cell communication inference for spatially resolved transcriptomic data
NotEnded99 / Fault Diagnosis Of Wheeled RobotFault Diagnosis of Wheeled Robot Based on Prior Knowledge and Spatial-temporal difference graph convolutional network
koujan / Robotics Course ProjectHaze can cause poor visibility and loss of contrast in images and videos. In this article, we study the dehazing problem which can improve visibility and thus help in many computer vision applications. An extensive comparison of state of the art single image dehazing methods is done. One simple contrast enhancement method is used for dehazing. Structure- texture decomposition has been used in conjunction with this enhancement method to improve its performance in presence of synthetic noise. Methods which use a haze formation model and attempt at solving an ill-posed problem using computer vision priors are also investigated. The two priors studied are dark channel prior and the non-local prior. Both qualitative and quantitative comparisons for atmospheric and underwater images on all three methods provide a conclusive idea of which dehazing method performs better. All this knowledge has been extended to video dehazing. A video dehazing method which uses the spatial and temporal information in a video is studied in depth. An improved version of video dehazing is proposed in this article, which uses the spatial-temporal information fusion framework but does not suffer from some of its limitations. The new video dehazing method is shown to produce better results on test videos
danielcamposramos / Knowledge3DWeb knowledge is fragmented — duplicated across fonts, embeddings, metadata, and renderings. Humans see pixels, AI sees tokens, neither shares the source. Knowledge3D: a sovereign GPU-native reference implementation for W3C PM-KR, where humans and AI consume the same procedural knowledge from one source.
MichaelBeechan / ThunderNet ReviewReal-time generic object detection on mobile platforms is a crucial but challenging computer vision task. However, previous CNN-based detectors suffer from enormous computational cost, which hinders them from real-time inference in computation-constrained scenarios. In this paper, we investigate the effectiveness of two-stage detectors in real-time generic detection and propose a lightweight twostage detector named ThunderNet. In the backbone part, we analyze the drawbacks in previous lightweight backbones and present a lightweight backbone designed for object detection. In the detection part, we exploit an extremely efficient RPN and detection head design. To generate more discriminative feature representation, we design two efficient architecture blocks, Context Enhancement Module and Spatial Attention Module. At last, we investigate the balance between the input resolution, the backbone, and the detection head. Compared with lightweight one-stage detectors, ThunderNet achieves superior performance with only 40% of the computational cost on PASCAL VOC and COCO benchmarks. Without bells and whistles, our model runs at 24.1 fps on an ARM-based device. To the best of our knowledge, this is the first real-time detector reported on ARM platforms. Code will be released for paper reproduction.
gengchenmai / Se KgeGengchen Mai, Krzysztof Janowicz, Ling Cai, Rui Zhu, Blake Regalia, Bo Yan, Meilin Shi, Ni Lao. SE-KGE: A Location-Aware Knowledge Graph Embedding Model for Geographic Question Answering and Spatial Semantic Lifting. Transactions in GIS. DOI:10.1111/TGIS.12629
yibie / Project NodalA local-first, infinite canvas that turns linear AI chats into a spatial knowledge graph.
HCPLab-SYSU / STKETSpatial-Temporal Knowledge-Embedded Transformer for Video Scene Graph Generation (TIP 2024, ACM MM 2023)
snousias / AvatreeThis paper presents AVATREE, a computational modelling framework that generates Anatomically Valid Airway tree conformations and provides capabilities for simulation of broncho-constriction apparent in obstructive pulmonary conditions. Such conformations are obtained from the personalized 3D geometry generated from computed tomography (CT) data through image segmentation. The patient-specific representation of the bronchial tree structure is extended beyond the visible airway generation depth using a knowledge-based technique built from morphometric studies. Additional functionalities of AVATREE include visualization of spatial probability maps for the airway generations projected on the CT imaging data, and visualization of the airway tree based on local structure properties. Furthermore, the proposed toolbox supports the simulation of broncho-constriction apparent in pulmonary diseases, such as chronic obstructive pulmonary disease (COPD) and asthma. AVATREE is provided as an open-source toolbox in C++ and is supported by a graphical user interface integrating the modelling functionalities. It can be exploited in studies of gas flow, gas mixing, ventilation patterns and particle deposition in the pulmonary system, with the aim to improve clinical decision making.
tjqansthd / Lap Rep KD DepthDecomposition and Replacement: Spatial Knowledge Distillation for Monocular Depth Estimation
wangdongjie100 / KDD2020Incremental Mobile User Profiling: Reinforcement Learning with Spatial Knowledge Graph for Modeling Event Streams
Zzzzz1 / CSKDOfficial code for Cumulative Spatial Knowledge Distillation for Vision Transformers (ICCV-2023) https://openaccess.thecvf.com/content/ICCV2023/html/Zhao_Cumulative_Spatial_Knowledge_Distillation_for_Vision_Transformers_ICCV_2023_paper.html
gcosne / OceanographyProjectToday satellites provide a surface signature of the temperature with a high spatial frequency: ie a good horizontal resolution but a low vertical resolution. Thanks to the ARGO database collected by buoys making vertical surveys, one has precise but sparse knowledge of the vertical thermal structure of the ocean. The objective of this project is to develop a methodology to statistically combine additional information to obtain a 3D time series sufficiently resolved horizontally and vertically to follow the eddies. It is based on a regression classification method that allows both to classify the temperature profiles and to propose a linear model between the satellite observations and the information of the buoys according to its label.
luilui97 / DSPPACCV2022 Source Code of paper "Feature Decoupled Knowledge Distillation via Spatial Pyramid Pooling"
FabriFalasca / Delta MAPSδ-MAPS is a method that identifies the semi-autonomous components of a spatio-temporal system and studies their weighted and potentially lagged interactions. The semi-autonomous components of the system are modeled as spatially contiguous, possibly overlapping, functionally homogeneous domains. To the best of our knowledge, δ-MAPS is the first method that can identify spatially contiguous and possibly overlapping clusters (i.e., domains). At a second step, δ-MAPS infers a directed and weighted network between the identified domains. Edge direction captures the temporal ordering of events while the weight associated to each edge captures the magnitude of interaction between domains.
tsinghua-fib-lab / Mobile Traffic Prediction Sigspatial23The official implementation of "Empowering Spatial Knowledge Graph for Mobile Traffic Prediction" (Sigspatial'23)
gouldju1 / Honr39900 Foundations Of Geospatial AnalyticsMaps are everywhere around us: in our cars, on our phones, and driving public health initiatives. Geospatial skills and knowledge are increasingly sought after in industry, and will continue to prove vital to Data Science. You will learn how to create maps and analyze spatial data using Python and SQL, how spatial data are applied in a variety of domains, and have hands-on experiences with real data. Together, we will answer questions such as: (1) what are maps, (2) how can we create maps from data, (3) and how do we quantify and analyze maps. Applied geospatial projects will include: autonomous vehicles, public health, supply chain, and more.
XuJin1992 / The Research And Implementation Of Data Mining For Geological DataData mining and knowledge discovery, refers to discover knowledge from huge amounts of data, has a broad application prospect.When faced with geological data, however, even the relatively mature existing models, there are defects performance and effect.Investigate its reason, mainly because of the inherent characteristics of geological data, high dimension, unstructured, more relevance, etc., in the data model, indexing structure knowledge representation, storage, mining, etc., is far more complicated than the traditional data. The geological data of the usual have raster, vector and so on, this paper pays attention to raster data processing.Tobler theorem tells us: geography everything associated with other things, but closer than far stronger correlation.Spatial correlation characteristics of geological data, the author of this paper, by establishing a spatial index R tree with spatial pattern mining algorithms as the guiding ideology, through the raster scanning method materialized space object space between adjacent relationship, transaction concept, thus the space with a pattern mining into the traditional association rules mining, and then take advantage of commonly used association rules to deal with some kind of geological data, to find association rules of interest. Using the simulation program to generate the geological data of the experiment, in the process of experiment, found a way to use R tree indexing can significantly speed up the generating spatial transaction set, at the same time, choose the more classic Apriori algorithm and FP - growth algorithm contrast performance, results show that the FP - growth algorithm is much faster than the Apriori algorithm, analyses the main reasons why the Apriori algorithm to generate a large number of candidate itemsets.In this paper, the main work is as follows: (1) In order to speed up the neighborhood search, choose to establish R tree spatial index, on the basis of summarizing the common scenarios to apply spatial indexing technology and the advantages and disadvantages. (2) Based on the analysis of traditional association rule mining algorithm and spatial association rule mining algorithm on the basis of the model based on event center space with pattern mining algorithm was described, and puts forward with a rule mining algorithm based on raster scanning, the algorithm by scanning for the center with a grid of R - neighborhood affairs set grid, will study data mining into the traditional data mining algorithm. (3) In the process of spatial index R tree insert, in order to prevent insertion to split after the leaf node, leading to a recursive has been split up destroy the one-way traverse, is put forward in the process of looking for insert position that records if full node number is M (M number) for each node up to insert nodes, first to divide to avoid after layers of recursive splitting up, speed up the R tree insertion efficiency. (4) On the basis of spatial transaction set preprocessing, realize the Apriori algorithm and FP-growth algorithm two kinds of classic association rule mining algorithm, performance contrast analysis.
zjunlp / Knowledge2Data[TASLP 2025] Spatial Knowledge Graph-Guided Synthesis for Multimodal LLMs
