UJI RobInLab's approach to the Amazon Robotics Challenge 2017

Pobil, Ángel P. del; Kassawat, Majd; Durán, Ángel; Arias, Marco; Nechyporenko, Nataliya; Mallick, Arijit; Cervera, E.; Subedi, Dipendra; Vasilev, Ilia; Cardin, D.; Sansebastiano, Emanuele; Martínez-Martín, Ester; Morales, Antonio; Casañ, Gustavo A.; Arenal, Ángel; Goriatcheff, B.; Rubert, Carlos; Recatala, G.

doi:10.1109/mfi.2017.8170448

Cited by 7 publications

(6 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to evaluate the importance of considering prior and external models in the design of deep network architectures, we propose three architectures to solve the depth estimation problem in a robot with an eye-in-hand camera for manipulation in an online shopping warehouse shelf ( Figure 1 ). In this scenario, fixed RGB-D cameras or laser sensors have been commonly used to get a faithful 3D representation of its surrounding space in order to deal with a large number of different items [ 32 ]. However, those sensors—either fixed or mounted on the robot—suffer from visibility issues to perceive objects such as those occluded or not visible within the shelf.…”

Section: Methodsmentioning

confidence: 99%

Integrating Sensor Models in Deep Learning Boosts Performance: Application to Monocular Depth Estimation in Warehouse Automation

Yoneyama

Durán

Pobil

2021

Sensors

View full text Add to dashboard Cite

Deep learning is the mainstream paradigm in computer vision and machine learning, but performance is usually not as good as expected when used for applications in robot vision. The problem is that robot sensing is inherently active, and often, relevant data is scarce for many application domains. This calls for novel deep learning approaches that can offer a good performance at a lower data consumption cost. We address here monocular depth estimation in warehouse automation with new methods and three different deep architectures. Our results suggest that the incorporation of sensor models and prior knowledge relative to robotic active vision, can consistently improve the results and learning performance from fewer than usual training samples, as compared to standard data-driven deep learning.

show abstract

Section: Methodsmentioning

confidence: 99%

Integrating Sensor Models in Deep Learning Boosts Performance: Application to Monocular Depth Estimation in Warehouse Automation

Yoneyama

Durán

Pobil

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…This architecture is mainly based on the VGG network [29] by reformulating the layers as learning residual functions. Given its good generalization performance, it was part of our implementation for the Amazon Picking Challenge in 2017 [30] (see Figure 8). However, the learning of a new object is too time consuming.…”

Section: Robots In Semi-structured Environmentsmentioning

confidence: 99%

Robot Vision for Manipulation: A Trip to Real-World Applications

Martínez-Martín

Pobil

2021

IEEE Access

View full text Add to dashboard Cite

“…This paper proposes a new approach to this problem based on the experience gathered by our participation in two editions of the Amazon Robotics Challenge. Indeed, a preliminary version of our system [2] was used in combination with an object recognition module [23] for successful participation in the Amazon Robotics Challenge 2017, (ARC'17) [24]. As a testbed for the experiments described in this paper, we use the scenario defined by that edition of the competition, which includes realistic constraints of an online shopping warehouse.…”

Section: Introductionmentioning

confidence: 99%

A Practical Approach for Picking Items in an Online Shopping Warehouse

et al. 2021

View full text Add to dashboard Cite

Commercially viable automated picking in unstructured environments by a robot arm remains a difficult challenge. The problem of robot grasp planning has long been around but the existing solutions tend to be limited when it comes to deploy them in open-ended realistic scenarios. Practical picking systems are called for that can handle the different properties of the objects to be manipulated, as well as the problems arising from occlusions and constrained accessibility. This paper presents a practical solution to the problem of robot picking in an online shopping warehouse by means of a novel approach that integrates a carefully selected method with a new strategy, the centroid normal approach (CNA), on a cost-effective dual-arm robotic system with two grippers specifically designed for this purpose: a two-finger gripper and a vacuum gripper. Objects identified in the scene point cloud are matched to the grasping techniques and grippers to maximize success. Extensive experimentation provides clues as to what are the reasons for success and failure. We chose as benchmark the scenario proposed by the 2017 Amazon Robotics Challenge, since it represents a realistic description of a retail shopping warehouse case; it includes many challenging constraints, such as a wide variety of different product items with a diversity of properties, which are also presented with restricted visibility and accessibility.

show abstract

UJI RobInLab's approach to the Amazon Robotics Challenge 2017

Cited by 7 publications

References 11 publications

Integrating Sensor Models in Deep Learning Boosts Performance: Application to Monocular Depth Estimation in Warehouse Automation

Integrating Sensor Models in Deep Learning Boosts Performance: Application to Monocular Depth Estimation in Warehouse Automation

Robot Vision for Manipulation: A Trip to Real-World Applications

A Practical Approach for Picking Items in an Online Shopping Warehouse

Contact Info

Product

Resources

About