Autonomous Terrain Classification With Co- and Self-Training Approach

Otsu, Kyohei; Ono, Masahiro; Fuchs, Thomas J.; Baldwin, Ian; Kubota, Takashi

doi:10.1109/lra.2016.2525040

Cited by 86 publications

(60 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is not sufficient for ground robots, which by definition interact with the ground and need to exert forces onto the environment to move forward. For semanticaware navigation, additional and richer sensor streams like grayscale [4], RGB [8], [5], [9], [10], [11], NIR [12], [5] and thermal cameras [13] as well as RADAR [14] have been used.…”

Section: B Related Workmentioning

confidence: 99%

Where Should I Walk? Predicting Terrain Properties From Images Via Self-Supervised Learning

Wellhausen

Dosovitskiy

Ranftl

et al. 2019

IEEE Robot. Autom. Lett.

157

120

View full text Add to dashboard Cite

Section: B Related Workmentioning

confidence: 99%

Where Should I Walk? Predicting Terrain Properties From Images Via Self-Supervised Learning

Wellhausen

Dosovitskiy

Ranftl

et al. 2019

IEEE Robot. Autom. Lett.

157

120

View full text Add to dashboard Cite

“…Semantic-aware navigation approaches typically leverage additional sensor modalities to infer additional terrain information [13], [14], [15], [16], [17], [18], [2], [19], [20]. Approaches using more unconventional sensors either require a long observation duration [19] or a bulky sensor payload [20] which exceeds the capabilities of our target platform.…”

Section: Related Workmentioning

confidence: 99%

Safe Robot Navigation Via Multi-Modal Anomaly Detection

Wellhausen

Ranftl

Hutter

2020

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Navigation in natural outdoor environments requires a robust and reliable traversability classification method to handle the plethora of situations a robot can encounter. Binary classification algorithms perform well in their native domain but tend to provide overconfident predictions when presented with out-of-distribution samples, which can lead to catastrophic failure when navigating unknown environments. We propose to overcome this issue by using anomaly detection on multi-modal images for traversability classification, which is easily scalable by training in a self-supervised fashion from robot experience. In this work, we evaluate multiple anomaly detection methods with a combination of uni-and multi-modal images in their performance on data from different environmental conditions. Our results show that an approach using a feature extractor and normalizing flow with an input of RGB, depth and surface normals performs best. It achieves over 95% area under the ROC curve and is robust to out-of-distribution samples.

show abstract

“…From the viewpoint of output as a measure of traversability for the self-supervised learning, the discrimination of traversable/non-traversable terrains [10,16] and the classification of terrains, such as gravel/sand/grass [19][20][21][22], are frequently used.…”

Section: Related Work On Unsupervised Learning Approaches To Traversmentioning

confidence: 99%

Regressed Terrain Traversability Cost for Autonomous Navigation Based on Image Textures

Bekhti

Kobayashi

2020

Applied Sciences

View full text Add to dashboard Cite

The exploration of remote, unknown, rough environments by autonomous robots strongly depends on the ability of the on-board system to build an accurate predictor of terrain traversability. Terrain traversability prediction can be made more cost efficient by using texture information of 2D images obtained by a monocular camera. In cases where the robot is required to operate on a variety of terrains, it is important to consider that terrains sometimes contain spiky objects that appear as non-uniform in the texture of terrain images. This paper presents an approach to estimate the terrain traversability cost based on terrain non-uniformity detection (TNUD). Terrain images undergo a multiscale analysis to determine whether a terrain is uniform or non-uniform. Terrains are represented using a texture and a motion feature computed from terrain images and acceleration signal, respectively. Both features are then combined to learn independent Gaussian Process (GP) predictors, and consequently, predict vibrations using only image texture features. The proposed approach outperforms conventional methods relying only on image features without utilizing TNUD.

show abstract

Autonomous Terrain Classification With Co- and Self-Training Approach

Cited by 86 publications

References 17 publications

Where Should I Walk? Predicting Terrain Properties From Images Via Self-Supervised Learning

Where Should I Walk? Predicting Terrain Properties From Images Via Self-Supervised Learning

Safe Robot Navigation Via Multi-Modal Anomaly Detection

Regressed Terrain Traversability Cost for Autonomous Navigation Based on Image Textures

Contact Info

Product

Resources

About