A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2024; you can also visit the original URL.
The file type is application/pdf
.
Filters
Trajectory Regularization Enhances Self-Supervised Geometric Representation
[article]
2024
arXiv
pre-print
We hope the proposed benchmark and methods offer new insights and improvements in self-supervised geometric representation learning. ...
On this benchmark, we study enhancing SSL geometric representations without sacrificing semantic classification accuracy. ...
We also propose methods to enhance the self-supervised geometric representation quality, including using mid-level representations and trajectory regularization. * Equal advising. ...
arXiv:2403.14973v1
fatcat:riliefjy6vee5h6y3wpmxkykmm
Enhancing motion trajectory segmentation of rigid bodies using a novel screw-based trajectory-shape representation
[article]
2023
arXiv
pre-print
This new representation is validated for a self-supervised segmentation approach, both in simulation and using real recordings of human-demonstrated pouring motions. ...
This representation consists of a geometric progress rate and a third-order trajectory-shape descriptor. ...
The main objective of this work is to enhance templatebased approaches in supervised and self-supervised trajectory segmentation by incorporating invariance with respect to time, coordinate frame, and ...
arXiv:2309.11413v1
fatcat:patdsbwuifdzla5uszt33frzvi
Sequential Adversarial Learning for Self-Supervised Deep Visual Odometry
[article]
2019
arXiv
pre-print
The updated representation is used for depth estimation. Besides, we tackle VO as a self-supervised image generation task and take advantage of Generative Adversarial Networks (GAN). ...
We propose a self-supervised learning framework for visual odometry (VO) that incorporates correlation of consecutive frames and takes advantage of adversarial learning. ...
Instead, the performance can be enhanced by taking geometric relations of sequential observations into account. ...
arXiv:1908.08704v1
fatcat:fz6aykzgsjft7lwjvsu5w7z3zi
Pre-training on Synthetic Driving Data for Trajectory Prediction
[article]
2023
arXiv
pre-print
Given the heavy reliance of current trajectory forecasting models on data-driven methodologies, we aim to tackle the challenge of learning general trajectory forecasting representations under limited data ...
(MAE) for trajectory forecasting. ...
Our data synthesis and self-supervised pre-training pipeline enhances prediction performance without extra real-world driving data. ...
arXiv:2309.10121v2
fatcat:2niezh2x65dhxpyml3axsny62q
Monocular Camera-Based Point-Goal Navigation by Learning Depth Channel and Cross-Modality Pyramid Fusion
2022
AAAI Conference on Artificial Intelligence
In the visual perception part, we firstly propose a Self-supervised Depth Estimation network (SDE) specially tailored for the monocular camera-based navigation agent. ...
In the navigation part, the extracted visual representations are fed to a navigation policy network to learn how to map the visual representations to agent actions effectively. ...
Here, we utilize three self-supervised losses in training. ...
dblp:conf/aaai/0002D0022
fatcat:2exxzqvuarf55iqi5ooj4yn4tq
Contrastive Learning for Graph-Based Vessel Trajectory Similarity Computation
2023
Journal of Marine Science and Engineering
A graph neural network encoder is used to extract spatial dependency from the trajectory graph to learn better trajectory representations. ...
trajectory graph. ...
Inspired by previous research on contrastive learning [26] , this module applies the self-supervised learning paradigm to learn representations of ship trajectories by leveraging the dissimilarity between ...
doi:10.3390/jmse11091840
fatcat:bnwzngz52nfg5leofjd7mfek5y
PreTraM: Self-Supervised Pre-training via Connecting Trajectory and Map
[article]
2022
arXiv
pre-print
In this paper, we propose PreTraM, a self-supervised pre-training scheme via connecting trajectories and maps for trajectory forecasting. ...
Contrastive Learning, where we enhance map representation with contrastive learning on large quantities of HD-maps. ...
Method We propose a novel self-supervised pre-training scheme by connecting trajectory and map (PreTraM) to enhance the trajectory and map representations when there are small-scale trajectory data, but ...
arXiv:2204.10435v1
fatcat:7ec3yeqq3nfq5dg5achevhhfki
Self-supervised Graph-based Point-of-interest Recommendation
[article]
2022
arXiv
pre-print
In order to counteract the scarcity and incompleteness of POI check-ins, we propose a novel self-supervised learning paradigm in \ssgrec, where the trajectory representations are contrastively learned ...
In particular, we devise a novel Graph-enhanced Self-attentive layer to incorporate the collaborative signals from both global transition graph and local trajectory graphs to uncover the transitional dependencies ...
and temporal effects from the constructed graphs; 4) self-supervised learning with augmented graphs to further enhance the expressiveness of the learned representations. ...
arXiv:2210.12506v1
fatcat:cjyu7l34pzfuvc2oy4ssh3pqfm
Collaborative Learning of Depth Estimation, Visual Odometry and Camera Relocalization from Monocular Videos
2020
Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence
Moreover, the Geometric Attention Guidance Model is introduced to exploit the geometric relevance among three branches during learning. ...
We argue that these two tasks should be strategically combined to leverage the complementary advantages, and be further improved by exploiting the 3D geometric information from depth data, which is also ...
These self-supervised methods tend to suffer from the violation of moving vehicles or people. ...
doi:10.24963/ijcai.2020/68
dblp:conf/ijcai/ZhaoBYT20
fatcat:pz5cw3qaqjhyvobynryqpla63a
Survey on Modeling of Articulated Objects
[article]
2024
arXiv
pre-print
Geometric representation. ...
Self-supervised learning methods. ...
arXiv:2403.14937v1
fatcat:bmdz6nmttnaznovnve5ogpsd6m
Deep Learning for Visual Localization and Mapping: A Survey
[article]
2023
arXiv
pre-print
Instead of creating hand-designed algorithms based on physical models or geometric theories, deep learning solutions provide an alternative to solve the problem in a data-driven way. ...
Benefiting from the ever-increasing volumes of data and computational power on devices, these learning methods are fast evolving into a new area that shows potentials to track self-motion and estimate ...
By leveraging self-supervised learning [16] , or reinforcement learning [231] - [233] , it would offer opportunities to self-update system (neural network) parameters, and be promising to enhancing ...
arXiv:2308.14039v1
fatcat:h26nrsakhbcm5ja7vhacld4t3u
A Survey of Data-Efficient Graph Learning
[article]
2024
arXiv
pre-print
Next, we systematically review recent advances on this topic from several key aspects, including self-supervised graph learning, semi-supervised graph learning, and few-shot graph learning. ...
To tackle this problem, tremendous efforts have been devoted to enhancing graph machine learning performance under low-resource settings by exploring various approaches to minimal supervision. ...
Self-supervised Graph Learn- ing (Self); Semi-supervised Graph Learning (Semi); Few-shot Graph Learning (Few-shot); Graph Representation Learning (GRL); Graph Clustering (GC); Classical Semi-supervised ...
arXiv:2402.00447v3
fatcat:zygm52p4rvhrraag46b7lyzzsy
A Comprehensive Survey of Data Augmentation in Visual Reinforcement Learning
[article]
2022
arXiv
pre-print
the actions to generate synthesized trajectories under a self-supervised cycle consistency constraint. ...
A naive approach to enhancing generalization is to apply regularization techniques originally developed for supervised learning [18, 19] , including 2 regularization [76] , entropy regularization [16 ...
arXiv:2210.04561v3
fatcat:u4sexzv37jfi7cqil47zg22bvy
Subspace Clustering for Action Recognition with Covariance Representations and Temporal Pruning
[article]
2020
arXiv
pre-print
To this end, we propose a novel subspace clustering method, which exploits covariance matrix to enhance the action's discriminability and a timestamp pruning approach that allow us to better handle the ...
Albeit state-of-the-art approaches designed for this application are all supervised, in this paper we pursue a more challenging direction: Solving the problem with unsupervised learning. ...
On the one hand, we encode the raw skeletal trajectories using a covariance representation, which has been shown to be effective for the solving HAR problems [10] . ...
arXiv:2006.11812v1
fatcat:d5pdt6dmlzehboxavtw4fmnz4e
Self-Supervised Monocular Depth and Ego-Motion Estimation in Endoscopy: Appearance Flow to the Rescue
[article]
2021
arXiv
pre-print
One widely adopted assumption of depth and ego-motion self-supervised learning is that the image brightness remains constant within nearby frames. ...
Recently, self-supervised learning technology has been applied to calculate depth and ego-motion from monocular videos, achieving remarkable performance in autonomous driving scenarios. ...
Self-Supervised Depth and Ego-Motion Estimation. ...
arXiv:2112.08122v1
fatcat:7yjhpa5o6vdvpcwasva3zqkeme
« Previous
Showing results 1 — 15 out of 5,702 results