An Introduction to Hyperdimensional Computing for Robotics

Neubert, Peer; Schubert, Stefan; Protzel, Peter

doi:10.1007/s13218-019-00623-z

Cited by 81 publications

(46 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…VSAs make use of additional operations on high-dimensional vectors. So far, VSAs have been applied in various fields including robotics [45], addressing catastrophic forgetting in deep neural networks [9], medical diagnosis [73], fault detection [29], analogy mapping [52], reinforcement learning [30], long-short term memory [11], text classification [31], and synthesis of finite state automata [49]. They have been used in combination with deep-learned descriptors before, e.g.…”

Section: Vector Symbolic Architecturesmentioning

confidence: 99%

“…They have been used in combination with deep-learned descriptors before, e.g. for sequence encoding [45] and local feature aggregation [44]. A particularly related VSA are spatial semantic pointers [34], a variant of the Semantic Pointer Architecture [15], that processes vector encodings of symbols with positions in images using a complex vector space and fractional binding [34].…”

Section: Vector Symbolic Architecturesmentioning

confidence: 99%

“…Although this type of computation is very common in the VSAs literature, it is significantly different to conventional algorithmic descriptions. Please refer to [26,45,51] for general introductions to Vector Symbolic Architectures.…”

Section: A Preleminaries On Vector Symbolic Architecturesmentioning

confidence: 99%

See 2 more Smart Citations

Vector Semantic Representations as Descriptors for Visual Place Recognition

Neubert¹,

Schubert²,

Schlegel³

et al. 2021

Robotics: Science and Systems XVII

Self Cite

View full text Add to dashboard Cite

Place recognition is the task of recognizing the current scene from a database of known places. The currently dominant algorithmic paradigm is to use (deep learning based) holistic feature vectors to describe each place and use fast vector query methods to find matchings. We propose a novel type of image descriptor, Vector Semantic Representations (VSR), that encodes the spatial semantic layout from a semantic segmentation together with appearance properties in a, for example, 4,096 dimensional vector for place recognition. We leverage operations from the established class of Vector Symbolic Architectures to combine symbolic (e.g. class label) and numeric (e.g. feature map response) information in a common vector representation. We evaluate the proposed semantic descriptor on 13 standard mobile robotic place recognition datasets and compare to six descriptors from the literature. VSR is on par with the best compared descriptor (NetVLAD) in terms of mean average precision and superior in terms of recall and worst-case average precision. This makes the approach particularly interesting for candidate selection. For a more detailed investigation, we discuss and evaluate recall integrity as additional criterion. Further, we demonstrate that the semantic descriptor is particularly well suited for combination with existing appearance descriptors indicating that semantics provide complementary information for image matching.

show abstract

Section: Vector Symbolic Architecturesmentioning

confidence: 99%

Section: Vector Symbolic Architecturesmentioning

confidence: 99%

See 1 more Smart Citation

Vector Semantic Representations as Descriptors for Visual Place Recognition

Neubert¹,

Schubert²,

Schlegel³

et al. 2021

Robotics: Science and Systems XVII

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the subsequent epochs iterations, HD updates the class hypervectors by observing if the model correctly predicts the training data. If the model mispredicts an encoded query H of label as class C ′ , HD updates as shown by Equation (9). If learning rate is not provided, SHEAR finds the best through bisectioning for a certain number of iterations.…”

Section: Software Layermentioning

confidence: 99%

“…This phase is commonly referred to in literature as "associative search". Despite the simplicity of this "learning" scheme, HD computing has been successfully applied to a number of practical problems in the literature ranging from optimizing the performance of web-browsers [5], to DNA sequence alignment [6], bio-signal processing [7], robotics [8,9], and privacy preserving learning [10,11]. The primary appeal of HD computing lies in its amenability to implementation in modern hardware accelerators.…”

Section: Introductionmentioning

confidence: 99%

SHEAR er

Khaleghi

Salamat

Thomas

et al. 2020

Proceedings of the ACM/IEEE International Symposium on Low Power Electronics and Design

View full text Add to dashboard Cite

Hyperdimensional computing (HD) is an emerging paradigm for machine learning based on the evidence that the brain computes on highdimensional, distributed, representations of data. The main operation of HD is encoding, which transfers the input data to hyperspace by mapping each input feature to a hypervector, followed by a bundling procedure that adds up the hypervectors to realize the encoding hypervector. The operations of HD are simple and highly parallelizable, but the large number of operations hampers the efficiency of HD in embedded domain. In this paper, we propose SHEAR , an algorithmhardware co-optimization to improve the performance and energy consumption of HD computing. We gain insight from a prudent scheme of approximating the hypervectors that, thanks to error resiliency of HD, has minimal impact on accuracy while provides high prospect for hardware optimization. Unlike previous works that generate the encoding hypervectors in full precision and then and then perform ex-post quantization, we compute the encoding hypervectors in an approximate manner that saves resources yet affords high accuracy. We also propose a novel FPGA architecture that achieves striking performance through massive parallelism with low power consumption. Moreover, we develop a software framework that enables training HD models by emulating the proposed approximate encodings. The FPGA implementation of SHEAR achieves an average throughput boost of 104,904× (15.7×) and energy savings of up to 56,044× (301×) compared to state-of-the-art encoding methods implemented on Raspberry Pi 3 (GeForce GTX 1080 Ti) using practical machine learning datasets.

show abstract

NetHD: Neurally Inspired Integration of Communication and Learning in Hyperspace

Poduval,

Ni,

Zou

et al. 2024

Advanced Intelligent Systems

View full text Add to dashboard Cite

The 6G network, the next‐generation communication system, is envisaged to provide unprecedented experience through hyperconnectivity involving everything. The communication should hold artificial intelligence‐centric network infrastructures as interconnecting a swarm of machines. However, existing network systems use orthogonal modulation and costly error correction code; they are very sensitive to noise and rely on many processing layers. These schemes impose significant overhead on low‐power internet of things devices connected to noisy networks. Herein, a hyperdimensional network‐based system, called , is proposed, which enables robust and efficient data communication/learning. exploits a redundant and holographic representation of hyperdimensional computing (HDC) to design highly robust data modulation, enabling two functionalities on transmitted data: 1) an iterative decoding method that translates the vector back to the original data without error correction mechanisms, or 2) a native hyperdimensional learning technique on transmitted data with no need for costly data decoding. A hardware accelerator that supports both data decoding and hyperdimensional learning using a unified accelerator is also developed. The evaluation shows that provides a bit error rate comparable to that of state‐of‐the‐art modulation schemes while achieving 9.4 faster and 27.8 higher energy efficiency compared to state‐of‐the‐art deep learning systems.

show abstract

An Introduction to Hyperdimensional Computing for Robotics

Cited by 81 publications

References 23 publications

Vector Semantic Representations as Descriptors for Visual Place Recognition

Vector Semantic Representations as Descriptors for Visual Place Recognition

SHEAR er

NetHD: Neurally Inspired Integration of Communication and Learning in Hyperspace

Contact Info

Product

Resources

About