computer-vision-1

Sequential Place Learning

3260 papers • 126 benchmarks • 313 datasets

State-of-the-art algorithms for route-based place recognition under changing conditions.

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in sequential-place-learning-2

Trend

Dataset

Best Model

Actions

No benchmarks available.

Libraries

i

Use these libraries to find sequential-place-learning-2 models and implementations

Datasets

No datasets available.

Subtasks

No subtasks available.

Most implemented papers

DeepSeqSLAM: A Trainable CNN+RNN for Joint Global Description and Sequence-based Place Recognition

Michael Milford, Marvin Chancán•Mon Nov 16 2020

DeepSeqSLAM is proposed: a trainable CNN+RNN architecture for jointly learning visual and positional representations from a single monocular image sequence of a route to further reduce false-positive rates compared to single-frame retrieval methods.

20

Content

0

Paper Graph

A Hybrid Compact Neural Architecture for Visual Place Recognition

Michael Milford, Marvin Chancán, Luis Hernandez-Nunez, A. Narendra, A. Barron•Mon Oct 14 2019

The resulting FlyNet+CANN network incorporates the compact pattern recognition capabilities of the FlyNet model with the powerful temporal filtering capabilities of an equally compact CANN, replicating entirely in a hybrid neural implementation the functionality that yields high performance in algorithmic localization approaches like SeqSLAM.

59 0

Paper Graph

SeqNet: Learning Descriptors for Sequence-Based Hierarchical Place Recognition

Michael Milford, Sourav Garg•Mon Feb 22 2021

A novel hybrid system is presented that creates a high performance initial match hypothesis generator using short learnt sequential descriptors, which enable selective control sequential score aggregation using single image learnt descriptors.

100 0

Paper Graph

Sequential Place Learning: Heuristic-Free High-Performance Long-Term Place Recognition

Michael Milford, Marvin Chancán•Mon Mar 01 2021

This work develops a joint visual and positional representation learning technique, via a sequential process, and designs a learning-based CNN+LSTM architecture, trainable via backpropagation through time, for viewpoint- and appearance-invariant place recognition.

6 0

Paper Graph

Adding a benchmark result helps the community track progress.