RGB Matters: Learning 7-DoF Grasp Poses on Monocular RGBD Images

Gou, Minghao; Fang, Hao-Shu; Zhu, Zhanda; Xu, Sheng; Wang, Chenxi; Lu, Cewu

doi:10.1109/icra48506.2021.9561409

Cited by 87 publications

(27 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Notably, the AP score of our method surpasses the others in all the test sets by a large margin. Even on the scenes with novel objects, the proposed model still has an averaged 5.0% improvement over the best baseline [41]. This implies that our model is able to generalize and perform well on novel objects.…”

Section: Resultsmentioning

confidence: 75%

“…Table I shows the performance of our approach compared to state of the art methods. We evaluated our trained model using the implementation of the evaluation metric shared by the authors of [11] enabling a direct comparison with the results of related works reported in [11], [41]. From the results presented in the table, we found that all the methods overall perform better in scenes with seen objects Fig.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Context-Aware Grasp Generation in Cluttered Scenes

Hoang

Stork

Stoyanov

2022

2022 International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

Conventional methods to autonomous grasping rely on a pre-computed database with known objects to synthesize grasps, which is not possible for novel objects. On the other hand, recently proposed deep learning-based approaches have demonstrated the ability to generalize grasp for unknown objects. However, grasp generation still remains a challenging problem, especially in cluttered environments under partial occlusion. In this work, we propose an end-to-end deep learning approach for generating 6-DOF collision-free grasps given a 3D scene point cloud. To build robustness to occlusion, the proposed model generates candidates by casting votes and accumulating evidence for feasible grasp configurations. We exploit contextual information by encoding the dependency of objects in the scene into features to boost the performance of grasp generation. The contextual information enables our model to increase the likelihood that the generated grasps are collision-free. Our experimental results confirm that the proposed system performs favorably in terms of predicting object grasps in cluttered environments in comparison to the current state of the art methods.

show abstract

Section: Resultsmentioning

confidence: 75%

Section: Resultsmentioning

confidence: 99%

Context-Aware Grasp Generation in Cluttered Scenes

Hoang

Stork

Stoyanov

2022

2022 International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

show abstract

“…A refinement step is also used for each grasp pose to allow further flexibility from the discretized coarse representation. Gou et al [96] estimate an SO(3) orientation and confidence of every pixel directly from an RGB image. An analytical method based on the depth image is then used to find the gripper width and position for each pixel.…”

Section: B Direct Regressionmentioning

confidence: 99%

“…(©2015 IEEE) custom datasets with those objects. A number of works used datasets collected using real robots [60,61,81,96,124] and some combined simulation and real-world data [61,87,96]. However, the large majority opted to use purely simulated datasets when training their networks.…”

Section: B Procedurally Generated Datasetsmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning Approaches to Grasp Synthesis: A Review

Newbury¹,

Gu²,

Lachlan³

et al. 2022

Preprint

View full text Add to dashboard Cite

Grasping is the process of picking an object by applying forces and torques at a set of contacts. Recent advances in deep-learning methods have allowed rapid progress in robotic object grasping. We systematically surveyed the publications over the last decade, with a particular interest in grasping an object using all 6 degrees of freedom of the end-effector pose. Our review found four common methodologies for robotic grasping: sampling-based approaches, direct regression, reinforcement learning, and exemplar approaches. Furthermore, we found two 'supporting methods' around grasping that use deeplearning to support the grasping process, shape approximation, and affordances. We have distilled the publications found in this systematic review (85 papers) into ten key takeaways we consider crucial for future robotic grasping and manipulation research.

show abstract

A Grasp Pose Detection Network Based on the DeepLabv3+ Semantic Segmentation Model

Zhang

2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

RGB Matters: Learning 7-DoF Grasp Poses on Monocular RGBD Images

Cited by 87 publications

References 32 publications

Context-Aware Grasp Generation in Cluttered Scenes

Context-Aware Grasp Generation in Cluttered Scenes

Deep Learning Approaches to Grasp Synthesis: A Review

A Grasp Pose Detection Network Based on the DeepLabv3+ Semantic Segmentation Model

Contact Info

Product

Resources

About