A Neurologically Inspired Sequence Processing Model for Mobile Robot Place Recognition

Neubert, Peer; Schubert, Stefan; Protzel, Peter

doi:10.1109/lra.2019.2927096

Cited by 50 publications

(49 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the section, we compare and show the results of our approach compared to the algorithm proposed in Neubert et al (2018) and Neubert et al (2019). In the experiments, a standard laptop with an i7-6500U CPU @ 2.50GHz was used.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…They discussed interesting theoretical association of aspects of HTM theory and the problem of mobile robot localization. Neubert et al (2019) extended the previous theoretical work and applied it to real world data, in particular, in combination with CNN-based image descriptors. The images were encoded with a CNN-based network like AlexNet (Krizhevsky et al, 2012) and NetVLAD (Arandjelović et al, 2018), and binarized with sparse Locality Sensitive Binary Hashing (sLSBH) based on random projections.…”

Section: Visual Place Recognitionmentioning

confidence: 91%

“…Differently from Neubert et al (2019), this paper implement a HTM algorithms for learning and prediction developed by Numenta (2018). The main advantage of our approach is scalability for the robot can run during long-term operations, as new minicolumns are not created during exploration of new areas.…”

Section: Hierarchical Temporal Memory Modelmentioning

confidence: 99%

“…The Figure 5 illustrates the overall approach and has similarities with the work proposed by Neubert et al (2019). Each image is processed by a visual front-end that extracts a descriptor.…”

Section: Nordland Datasetmentioning

confidence: 99%

“…MCN creates an internal representation that encodes sequential context through binary sparse distributed representations. This approach was extended in Neubert et al (2019), where each image is processed by CNN-based descriptors. Their encoding is a sparsified binary adaptation of locality sensitive hashing (LSH) based on random projections.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Long-Term Visual Robot Localization using Computational Models of the Neocortex

Pizzino

Vargas²,

Costa

2020

Anais Do Congresso Brasileiro De Automática 2020

View full text Add to dashboard Cite

Visual place recognition is an essential capability for autonomous mobile robots which use cameras as their primary sensors. Although there has been a considerable amount of research in the topic, the high degree of image variability poses extra research challenges. Following advances in neuroscience, new biologically inspired models have been developed. Inspired by the human neocortex, hierarchical temporal memory model has potential to identify temporal sequences of spatial patterns using sparse distributed representations, which are known to have high representational capacity and high tolerance to noise. These features are interesting for place recognition applications. Some authors have proposed simplifications from the original framework, such as starting from an empty set of minicolumns and increasing the number of minicolumns on demand instead of the usage of a fixed number of minicolumns whose connections adapt over time. In this paper, we investigate the usage of framework originally proposed with the aim of extending the run-time during long-term operations. Results show that the proposed architecture can encode an internal representation of the world using a fixed number of cells in order to improve system scalability.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%