VAST: Visual and Spectral Terrain Classification in Unstructured Multi-Class Environments

Hanson, Nathaniel; Shaham, Michael; Erdoğmuş, Deni̇z; Padır, Taşkın

doi:10.1109/iros47612.2022.9982078

Cited by 9 publications

(11 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sensory gaps in material perception are especially apparent where the deformation of surfaces might cause a failed grasp, or stuck vehicle. In many works, including several not included in the scope of this dissertation [200,201,199], we explored the use of spectroscopy to make an impact in many challenging robot perception problems. We successfully deployed systems that can estimate material class and regress physical quantities in the real world.…”

Section: Discussionmentioning

confidence: 99%

“…This sensor setup enables robots to first detect what the surface is, so they can plan how to act on them. This work was originally published in [145].…”

Section: Local Terrain Sensingmentioning

confidence: 99%

“…Our approach to this problem of differentiating intra-class variations builds on our previous work in terrain classification with point-based spectroscopy and multimodal sensing [145].…”

Section: Architecture and Calibrationmentioning

confidence: 99%

“…As an additional demonstration of the spectral acquisition capabilities of the sensor, we curate a collection of household items. This list includes visually confusing samples similar to the analysis we conducted in [199], such as real apples and faux apples. We also include baby oil (mineral oil), water, and isopropyl alcohol inside glass bottles to evaluate the system performance on visually clear items similar to the dataset from [178].…”

Section: Sample Spectral Measurementsmentioning

confidence: 99%

“…In knowing how spectral reflectance varies with distance offset and surface normal, robots will be equipped to make inferences based on spectroscopy on the fly, with confidence bounds on whether the current signals are actionable. Our preliminary work in this domain shows the recursive estimation of spectral probabilities is an appropriate way to encode confidence [199].…”

Section: Quantifying Distance Uncertainty With Spectroscopymentioning

confidence: 99%

See 4 more Smart Citations

Material informed robotics - spectral perception for object identification and parameter inference

Hanson

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

“…This sensor setup enables robots to first detect what the surface is, so they can plan how to act on them. This work was originally published in [145].…”

Section: Local Terrain Sensingmentioning

confidence: 99%

“…Our approach to this problem of differentiating intra-class variations builds on our previous work in terrain classification with point-based spectroscopy and multimodal sensing [145].…”

Section: Architecture and Calibrationmentioning

confidence: 99%

Section: Sample Spectral Measurementsmentioning

confidence: 99%

Section: Quantifying Distance Uncertainty With Spectroscopymentioning

confidence: 99%

See 3 more Smart Citations

Material informed robotics - spectral perception for object identification and parameter inference

Hanson

View full text Add to dashboard Cite

show abstract

Terrain‐aware path planning via semantic segmentation and uncertainty rejection filter with adversarial noise for mobile robots

Lee,

Lee

2024

Journal of Field Robotics

View full text Add to dashboard Cite

In ground mobile robots, effective path planning relies on their ability to assess the types and conditions of the surrounding terrains. Neural network‐based methods, which primarily use visual images for terrain classification, are commonly employed for this purpose. However, the reliability of these models can vary due to inherent discrepancies between the training images and the actual environment, leading to erroneous classifications and operational failures. Retraining models with additional images from the actual operating environment may enhance performance, but obtaining these images is often impractical or impossible. Moreover, retraining requires substantial offline processing, which cannot be performed online by the robot within an embedded processor. To address this issue, this paper proposes a neural network‐based terrain classification model, trained using an existing data set, with a novel uncertainty rejection filter (URF) for terrain‐aware path planning of mobile robots operating in unknown environments. A robot, equipped with a pretrained model, initially collects a small number of images (10 in this work) from its current environment to set the target uncertainty ratio of the URF. The URF then dynamically adjusts its sensitivity parameters to identify uncertain regions and assign associated traversal costs. This process occurs entirely online, without the need for offline procedures. The presented method was evaluated through simulations and physical experiments, comparing the point‐to‐point trajectories of a mobile robot equipped with (1) the neural network‐based terrain classification model combined with the presented adaptive URF, (2) the classification model without the URF, and (3) the classification model combined with a nonadaptive version of the URF. Path planning performance measured the Hausdorff distances between the desired and actual trajectories and revealed that the adaptive URF significantly improved performance in both simulations and physical experiments (conducted 10 times for each setting). Statistical analysis via t‐tests confirmed the significance of these results.

show abstract

Camera-Lidar-Based Terrain Multi-type Classification Using Both Spatial and Histogram Features of Lidars

Lu,

Wang,

et al. 2023

2023 3rd International Conference on Computer, Control and Robotics (ICCCR)

View full text Add to dashboard Cite

VAST: Visual and Spectral Terrain Classification in Unstructured Multi-Class Environments

Cited by 9 publications

References 32 publications

Material informed robotics - spectral perception for object identification and parameter inference

Material informed robotics - spectral perception for object identification and parameter inference

Terrain‐aware path planning via semantic segmentation and uncertainty rejection filter with adversarial noise for mobile robots

Camera-Lidar-Based Terrain Multi-type Classification Using Both Spatial and Histogram Features of Lidars

Contact Info

Product

Resources

About