Filters








10,936 Hits in 6.1 sec

Causal Navigation by Continuous-time Neural Networks [article]

Charles Vorbach, Ramin Hasani, Alexander Amini, Mathias Lechner, Daniela Rus
2021 arXiv   pre-print
In this paper, we propose a theoretical and experimental framework for learning causal representations using continuous-time neural networks, specifically over their discrete-time counterparts.  ...  Our results demonstrate that causal continuous-time deep models can perform robust navigation tasks, where advanced recurrent models fail.  ...  Let f be the nonlinearity of a continuous-time neural network.  ... 
arXiv:2106.08314v2 fatcat:bz6kyrwunjcgnflaeqm5wgv5vm

Causal networks in simulated neural systems

Anil K. Seth
2007 Cognitive Neurodynamics  
Analysis of a simple targetfixation model shows that causal networks provide intuitive representations of neural dynamics during behavior which can be validated by lesion experiments.  ...  Neural systems can therefore be analyzed in terms of causal networks, without assumptions about information processing, neural coding, and the like.  ...  networks were modified during learning of a spatial navigation task.  ... 
doi:10.1007/s11571-007-9031-z pmid:19003473 pmcid:PMC2289248 fatcat:erlleepwuzejdhzq2xeztfhvxy

Computational Methods in Social Neuroscience: Recent Advances, New Tools, and Future Directions

Carolyn Parkinson
2021 Social Cognitive and Affective Neuroscience  
neural response patterns and psychological and behavioral phenomena, examining time-varying patterns of connectivity between brain regions, and characterizing the social networks in which social thought  ...  These include approaches for modeling social decisions, characterizing multivariate neural response patterns at varying spatial scales, using decoded neurofeedback to draw causal links between specific  ...  Continuing to integrate approaches from ethology and behavioral ecology into social neuroscience promises to enrich our understanding of the neural mechanisms underlying social thought and behavior by  ... 
doi:10.1093/scan/nsab073 pmid:34101815 pmcid:PMC8343570 fatcat:4bmn565chzfznlvl7zopvw2rea

Deep Learning based Pedestrian Inertial Navigation: Methods, Dataset and On-Device Inference [article]

Changhao Chen, Peijun Zhao, Chris Xiaoxuan Lu, Wei Wang, Andrew Markham, Niki Trigoni
2020 arXiv   pre-print
Recently, there has been a growing interest in applying deep neural networks (DNNs) to motion sensing and location estimation.  ...  Exploiting inertial data for accurate and reliable pedestrian navigation supports is a key component for emerging Internet-of-Things applications and services.  ...  To enhance the online efficiency of DNN models on mobile devices, we propose Light Inertial Odometry Neural Networks (L-IONet), a lightweight deep neural network framework to learn inertial navigation  ... 
arXiv:2001.04061v1 fatcat:6vzynfiklza7la4dnejyodoyky

Hierarchical RNNs-Based Transformers MADDPG for Mixed Cooperative-Competitive Environments [article]

Xiaolong Wei, LiFang Yang, Xianglin Huang, Gang Cao, Tao Zhulin, Zhengyang Du, Jing An
2021 arXiv   pre-print
correlations between multiple sequences so that we can capture the causal relationship between sub-time sequences and make HRTMADDPG more efficient.  ...  By establishing dynamically weighted parameters for choosing relevant and irrelevant features, the key information can be strengthened, and the irrelevant information can be weakened.  ...  FLOATER is a recursive model, and functions can be modeled by neural networks.  ... 
arXiv:2105.04888v1 fatcat:m6bciz74bvh7vkw6fyzdes6mwe

Granger causal connectivity dissociates navigation networks that subserve allocentric and egocentric path integration

Chin-Teng Lin, Te-Cheng Chiu, Yu-Kai Wang, Chun-Hsiang Chuang, Klaus Gramann
2018 Brain Research  
how different areas of the navigational network interact, 17 we investigated the dynamic causal interactions of brain regions involved in 18 solving a virtual navigation task.  ...  EEG signals were decomposed by independent 19 component analysis (ICA) and subsequently examined for information flow 20 between clusters of independent components (ICs) using direct short-time 21 directed  ...  cortices (see 251 Posterior navigation (RSC-related) network 293 In addition to the anterior navigation network, a posterior navigation 294 network revealed the RSC to be causally connected with the  ... 
doi:10.1016/j.brainres.2017.11.016 pmid:29158177 fatcat:iezjgvwljvd2vdqwn6bjk3yzba

Understanding images in biological and computer vision

Andrew J. Schofield, Iain D. Gilchrist, Marina Bloj, Ales Leonardis, Nicola Bellotto
2018 Interface Focus  
activation of the other, this causal relationship being observed by a third neuron [21] .  ...  Continuing the neural network theme, Jitendra Malik presented a review of object recognition methods in machine vision including DCNNs and outlined the ways in which such systems still fall short of biological  ... 
doi:10.1098/rsfs.2018.0027 fatcat:xi7fvjkzozgj3pekdvh6kbcx44

Artificial Neural Network Algorithm for Online Glucose Prediction from Continuous Glucose Monitoring

C. Pérez-Gandía, A. Facchinetti, G. Sparacino, C. Cobelli, E.J. Gómez, M. Rigla, A. de Leiva, M.E. Hernando
2010 Diabetes Technology & Therapeutics  
Methods: The predictor is implemented with an artificial neural network model (NNM).  ...  and Aims: Continuous glucose monitoring (CGM) devices could be useful for real-time management of diabetes therapy.  ...  Before feeding the prediction algorithm, it was found convenient to reduce the effect of noise in the FreeStyle Navigator® profiles (1 min. sampling) by pre-filtering them using a causal Kalman filtering  ... 
doi:10.1089/dia.2009.0076 pmid:20082589 fatcat:l72qsy4ysvamvcxkjblblllhbu

Explainable Deep Reinforcement Learning for UAV Autonomous Navigation [article]

Lei He, Aouf Nabil, Bifeng Song
2021 arXiv   pre-print
Moreover, some global analysis are also provided for experts to evaluate and improve the trained neural network.  ...  In this paper, a neural network-based reactive controller is proposed for a quadrotor to fly autonomously in unknown outdoor environment.  ...  In this paper, an end-to-end neural network is proposed to address the UAV reactive navigation problem in the complex unknown environment for small UAVs with SWaP constraints.  ... 
arXiv:2009.14551v2 fatcat:u76wnblsgneszo3i7tnw6wqzhq

Machine Learning Incorporated With Causal Analysis for Short-Term Prediction of Sea Ice

Ming Li, Ren Zhang, Kefeng Liu
2021 Frontiers in Marine Science  
The experimental results in the Barents-Kara (B-K) sea show: (1) compared with correlation analysis, the input variables of ML models screened out by causal analysis achieve better prediction; (2) when  ...  Accurate and fast prediction of sea ice conditions is the foundation of safety guarantee for Arctic navigation.  ...  continuous time series prediction.  ... 
doi:10.3389/fmars.2021.649378 fatcat:7be3olyftnbabl4lwpcxsg3h7u

Learning interaction rules from multi-animal trajectories via augmented behavioral models [article]

Keisuke Fujii, Naoya Takeishi, Kazushi Tsutsui, Emyo Fujioka, Nozomi Nishiumi, Ryoya Tanaka, Mika Fukushiro, Kaoru Ide, Hiroyoshi Kohno, Ken Yoda, Susumu Takahashi, Shizuko Hiryu (+1 others)
2021 arXiv   pre-print
We adopt an approach for augmenting incomplete multi-agent behavioral models described by time-varying dynamical systems with neural networks.  ...  This can provide interpretable signs of Granger-causal effects over time, i.e., when specific others cause the approach or separation.  ...  Acknowledgments This work was supported by JSPS KAKENHI (Grant Numbers 19H04941, 20H04075, 16H06541, 25281056, 19H01131, 18H03786, 16H06542, 19H04939, and JP18H03287), JST PRESTO (JP-MJPR20CA), and JST  ... 
arXiv:2107.05326v3 fatcat:yxj3vbgmivc4jggw5cersb2mnm

The Head-Direction Signal Plays a Functional Role as a Neural Compass during Navigation

William N. Butler, Kyle S. Smith, Matthijs A.A. van der Meer, Jeffrey S. Taube
2017 Current Biology  
These results provide evidence that the HD signal plays a causal role as a neural compass in navigation. ‡ RESULTS We injected a retrogradely transported canine adenovirus carrying Cre recombinase (CAV2  ...  Previous work has found that HD cell tuning correlates with behavior on navigational tasks, but a direct, causal link between HD cells and navigation has not been demonstrated.  ...  our interpretation that these shifts in navigational behavior were caused by the animals' reliance on an unstable neural compass.  ... 
doi:10.1016/j.cub.2017.03.033 pmid:28416119 pmcid:PMC5425164 fatcat:o63d5kduejcrxceypv3dpvn27m

From semantics to execution: Integrating action planning with reinforcement learning for robotic causal problem-solving [article]

Manfred Eppe, Phuong D.H. Nguyen, Stefan Wermter
2019 arXiv   pre-print
A problem with the integration of both approaches is that action planning is based on discrete high-level action- and state spaces, whereas reinforcement learning is usually driven by a continuous reward  ...  Symbolic action planning is useful to resolve causal dependencies and to break a causally complex problem down into a sequence of simpler high-level actions.  ...  For the real application, the physical training time can potentially further be lowered by applying more neural network training batches per rollout, and by performing pre-training using the simulation  ... 
arXiv:1905.09683v2 fatcat:zoqszldm6rdt7a4jntt3fbjuuq

Training a Resilient Q-network against Observational Interference

Chao-Han Huck Yang, I-Te Danny Hung, Yi Ouyang, Pin-Yu Chen
2022 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Inspired by causal inference for observational interference, we propose a causal inference based DQN algorithm called causal inference Q-network (CIQ).  ...  In practice, however, a DRL agent may receive faulty observation by abrupt interferences such as black-out, frozen-screen, and adversarial perturbation.  ...  During training, switching between the two neural networks is determined by the training interference label i train t .  ... 
doi:10.1609/aaai.v36i8.20862 fatcat:st3hmlrx75a5xdy37tusxlglo4

From Semantics to Execution: Integrating Action Planning With Reinforcement Learning for Robotic Causal Problem-Solving

Manfred Eppe, Phuong D H Nguyen, Stefan Wermter
2019 Frontiers in Robotics and AI  
A problem with the integration of both approaches is that action planning is based on discrete high-level action- and state spaces, whereas reinforcement learning is usually driven by a continuous reward  ...  Symbolic action planning is useful to resolve causal dependencies and to break a causally complex problem down into a sequence of simpler high-level actions.  ...  For the real application, the physical training time can potentially further be lowered by applying more neural network training batches per rollout, and by performing pre-training using the simulation  ... 
doi:10.3389/frobt.2019.00123 pmid:33501138 pmcid:PMC7805615 fatcat:zfuealfbgfce7jedo5e27yvtsu
« Previous Showing results 1 — 15 out of 10,936 results