Robot learning through observation via coarse-to-fine grained video summarization

The West African Craton (WAC) is one of the major cratons in the Rodinia jigsaw puzzle (~1000&#8211;750 Ma). In the Rodinian models, the position of West Africa is mainly constrained by the assumption that it had been a partner of Amazonia since the Paleoproterozoic. Unfortunately, no paleomagnetic data are available for these cratons when the Rodina supercontinent is considered tectonically stable (~1000-750 Ma). Thus, every new reliable paleomagnetic pole for the West African Craton during the Neoproterozoic times is of paramount importance to constrain its position and testing the Rodinia models. In this study we present a combined paleomagnetic and geochronological investigation for the Manso dyke swarm in the Leo-Man Shield, southern West Africa (Ghana). The ~860 Ma emplacement age for the NNW-trending Manso dykes is thus well-constrained by two new U-Pb apatite ages of 857.2 &#177; 8.5 Ma and 855 &#177; 16 Ma, in agreement with baddeleyite data. Remanence of these coarse-to-fine grained dolerite dykes is carried by stable single to pseudo-single domain (SD-PSD) magnetite. A positive baked-contact test, associated to a positive reversal test (Class-C), support the primary remanence obtained for these dykes (13 sites). Moreover, our new paleomagnetic dataset satisfy all the seven R-criteria (R=7). The ~860 Ma Manso pole can thus be considered as the first key Tonian paleomagnetic pole for West Africa. We propose that the West Africa-Baltica-Amazonia-Congo-S&#227;o Francisco were associated in a long-lived WABAMGO juxtaposition (~1100&#8211;800 Ma).Keywords: West Africa, Neoproterozoic, Tonian, Rodinia, paleomagnetism.&#160;

Download Full-text

Generalized Zero-Shot Vehicle Detection in Remote Sensing Imagery via Coarse-to-Fine Framework

Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence ◽

10.24963/ijcai.2019/97 ◽

2019 ◽

Author(s):

Hong Chen ◽

Yongtan Luo ◽

Liujuan Cao ◽

Baochang Zhang ◽

Guodong Guo ◽

...

Keyword(s):

Remote Sensing ◽

Vehicle Detection ◽

Training Data ◽

Coarse Grained ◽

Vehicle Classification ◽

Vehicle Localization ◽

Fine Grained ◽

Coarse To Fine ◽

Generalized Zero ◽

Detection And Recognition

Vehicle detection and recognition in remote sensing images are challenging, especially when only limited training data are available to accommodate various target categories. In this paper, we introduce a novel coarse-to-fine framework, which decomposes vehicle detection into segmentation-based vehicle localization and generalized zero-shot vehicle classification. Particularly, the proposed framework can well handle the problem of generalized zero-shot vehicle detection, which is challenging due to the requirement of recognizing vehicles that are even unseen during training. Specifically, a hierarchical DeepLab v3 model is proposed in the framework, which fully exploits fine-grained features to locate the target on a pixel-wise level, then recognizes vehicles in a coarse-grained manner. Additionally, the hierarchical DeepLab v3 model is beneficially compatible to combine the generalized zero-shot recognition. To the best of our knowledge, there is no publically available dataset to test comparative methods, we therefore construct a new dataset to fill this gap of evaluation. The experimental results show that the proposed framework yields promising results on the imperative yet difficult task of zero-shot vehicle detection and recognition.

Download Full-text

From Coarse to Fine-Grained Parcellation of the Cortical Surface Using a Fiber-Bundle Atlas

Frontiers in Neuroinformatics ◽

10.3389/fninf.2020.00032 ◽

2020 ◽

Vol 14 ◽

Author(s):

Narciso López-López ◽

Andrea Vázquez ◽

Josselin Houenou ◽

Cyril Poupon ◽

Jean-François Mangin ◽

...

Keyword(s):

Fiber Bundle ◽

Cortical Surface ◽

Fine Grained ◽

Coarse To Fine

Download Full-text

Mirroring without Overimitation: Learning Functionally Equivalent Manipulation Actions

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v33i01.33018025 ◽

2019 ◽

Vol 33 ◽

pp. 8025-8033 ◽

Cited By ~ 1

Author(s):

Hangxin Liu ◽

Chi Zhang ◽

Yixin Zhu ◽

Chenfanfu Jiang ◽

Song-Chun Zhu

Keyword(s):

Mirror Neurons ◽

Haptic Feedback ◽

Robot Learning ◽

Learning From Demonstration ◽

Fine Grained ◽

Model Free ◽

Simulation Engine ◽

State Changes ◽

Manipulation Task ◽

Physics Based Simulation

This paper presents a mirroring approach, inspired by the neuroscience discovery of the mirror neurons, to transfer demonstrated manipulation actions to robots. Designed to address the different embodiments between a human (demonstrator) and a robot, this approach extends the classic robot Learning from Demonstration (LfD) in the following aspects:i) It incorporates fine-grained hand forces collected by a tactile glove in demonstration to learn robot’s fine manipulative actions; ii) Through model-free reinforcement learning and grammar induction, the demonstration is represented by a goal-oriented grammar consisting of goal states and the corresponding forces to reach the states, independent of robot embodiments; iii) A physics-based simulation engine is applied to emulate various robot actions and mirrors the actions that are functionally equivalent to the human’s in the sense of causing the same state changes by exerting similar forces. Through this approach, a robot reasons about which forces to exert and what goals to achieve to generate actions (i.e., mirroring), rather than strictly mimicking demonstration (i.e., overimitation). Thus the embodiment difference between a human and a robot is naturally overcome. In the experiment, we demonstrate the proposed approach by teaching a real Baxter robot with a complex manipulation task involving haptic feedback—opening medicine bottles.

Download Full-text

A Coarse-to-Fine Adaptive Network for Appearance-Based Gaze Estimation

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v34i07.6636 ◽

2020 ◽

Vol 34 (07) ◽

pp. 10623-10630 ◽

Cited By ~ 1

Author(s):

Yihua Cheng ◽

Shiyao Huang ◽

Fei Wang ◽

Chen Qian ◽

Feng Lu

Keyword(s):

State Of The Art ◽

Face Image ◽

Gaze Direction ◽

Gaze Estimation ◽

Adaptive Network ◽

Fine Grained ◽

Intrinsic Correlation ◽

Parallel Feature ◽

Coarse To Fine

Human gaze is essential for various appealing applications. Aiming at more accurate gaze estimation, a series of recent works propose to utilize face and eye images simultaneously. Nevertheless, face and eye images only serve as independent or parallel feature sources in those works, the intrinsic correlation between their features is overlooked. In this paper we make the following contributions: 1) We propose a coarse-to-fine strategy which estimates a basic gaze direction from face image and refines it with corresponding residual predicted from eye images. 2) Guided by the proposed strategy, we design a framework which introduces a bi-gram model to bridge gaze residual and basic gaze direction, and an attention component to adaptively acquire suitable fine-grained feature. 3) Integrating the above innovations, we construct a coarse-to-fine adaptive network named CA-Net and achieve state-of-the-art performances on MPIIGaze and EyeDiap.

Download Full-text