Goto

Collaborating Authors

Object-Oriented Architecture


15 of the best online Georgia Tech courses you can take for free

Mashable

These free courses do not come with a certificate of completion, but that's the only catch. You can still learn at your own pace with unrestricted access to all the course materials. Find the best free online courses from Georgia Tech on edX.


Learn Python for less than $25 through October 15th

PCWorld

Python is the world's most popular programming language not just because it's relatively easy to learn, but because it's used in so many applications and it's highly scalable. If you've ever wanted to learn to code, starting with Python is a great idea. And it's an even better idea now because during Deal Days, you can get The Premium Python Programming Certification Bundle for $23.97 (reg. This bundle includes ten courses from top instructors like Joe Rahl (4.6/5-star instructor rating) and Edouard Renard (4.6/5-star rating). Starting with the absolute basics of Python, you'll learn basic coding principles, explore Object-Oriented Programming (OOP), and much more as you slowly level up your skills.


15 of the best Georgia Tech courses you can take online for free

Mashable

These free courses do not include a certificate of completion, so if you really want something to stick on your CV, you will need to pay a small fee. This is Georgia Tech we're talking about, so it might be worth it. Find the best free Georgia Tech courses on edX.


1MDB suspect and Jho Low associate dies weeks after questioning

Al Jazeera

A suspect in the 1MDB scandal has died weeks after being deported to Malaysia to face questioning over his role in the $4.5bn fraud. Kee Kok Thiam died in hospital on Monday following a "sudden massive stroke" and was cremated on Wednesday morning, Kee's family said in a statement. "We urge all parties not to entertain any speculations on this unfortunate event and allow the family the space to grief [sic] on his passing," the statement said. News of the 56-year-old businessman's death comes hours after Al Jazeera reported that the Malaysian Anti-Corruption Commission (MACC) had confirmed the whereabouts of fugitive Malaysian financier Jho Taek Low – the alleged mastermind of the 1MDB scandal – in Macau based on its questioning of Kee. The MACC said that Kee, who was deported from Macau earlier this month, revealed he had met with Low and other 1MDB fugitives in the Chinese territory and that Low had instructed him "not to return to Malaysia as a witness in the 1MDB case".


3D Object Proposals for Accurate Object Class Detection Xiaozhi Chen 1 Andrew Berneshawi

Neural Information Processing Systems

The goal of this paper is to generate high-quality 3D object proposals in the context of autonomous driving. Our method exploits stereo imagery to place proposals in the form of 3D bounding boxes. We formulate the problem as minimizing an energy function encoding object size priors, ground plane as well as several depth informed features that reason about free space, point cloud densities and distance to the ground. Our experiments show significant performance gains over existing RGB and RGB-D object proposal methods on the challenging KITTI benchmark. Combined with convolutional neural net (CNN) scoring, our approach outperforms all existing results on all three KITTI object classes.


Hierarchical Object Representation for Open-Ended Object Category Learning and Recognition

Neural Information Processing Systems

Most robots lack the ability to learn new objects from past experiences. To migrate a robot to a new environment one must often completely re-generate the knowledgebase that it is running with. Since in open-ended domains the set of categories to be learned is not predefined, it is not feasible to assume that one can pre-program all object categories required by robots. Therefore, autonomous robots must have the ability to continuously execute learning and recognition in a concurrent and interleaved fashion. This paper proposes an open-ended 3D object recognition system which concurrently learns both the object categories and the statistical features for encoding objects. In particular, we propose an extension of Latent Dirichlet Allocation to learn structural semantic features (i.e.


Unsupervised learning of object frames by dense equivariant image labelling James Thewlis

Neural Information Processing Systems

One of the key challenges of visual perception is to extract abstract models of 3D objects and object categories from visual measurements, which are affected by complex nuisance factors such as viewpoint, occlusion, motion, and deformations. Starting from the recent idea of viewpoint factorization, we propose a new approach that, given a large number of images of an object and no other supervision, can extract a dense object-centric coordinate frame. This coordinate frame is invariant to deformations of the images and comes with a dense equivariant labelling neural network that can map image pixels to their corresponding object coordinates. We demonstrate the applicability of this method to simple articulated objects and deformable objects such as human faces, learning embeddings from random synthetic transformations or optical flow correspondences, all without any manual supervision.


Reviews: Zero-Shot Transfer with Deictic Object-Oriented Representation in Reinforcement Learning

Neural Information Processing Systems

Post rebuttal: I now understand the middle ground this paper is positioned, and the difference to propositional OO representations where you don't necessarily care which instance of an object type you're dealing with, which significantly reduces the dimensionality of learning transition dynamics. But this is still similar to other work on graph neural networks for model learning in fully relational representations, like Relation Networks by Santoro et al., and Interaction Networks by Battaglia et al. which in worst case learn T * n * (n-1) relations for n objects for T types of relations. However, this paper does do a nice job of formalizing from the OO-MDP and Propositional MDP setting as opposed to the two papers I mentioned which do not, and focus on the physical dynamics case. I am willing to increase my score based on this, but still do not think it is novel enough to be accepted. This is very similar to relational MDPs, but they learn transition dynamics in this relational attribute space rather than real state space.


Zero Shot Transfer with Deictic Object Oriented Representation in Reinforcement Learning

Neural Information Processing Systems

Object-oriented representations in reinforcement learning have shown promise in transfer learning, with previous research introducing a propositional objectoriented framework that has provably efficient learning bounds with respect to sample complexity. However, this framework has limitations in terms of the classes of tasks it can efficiently learn. In this paper we introduce a novel deictic objectoriented framework that has provably efficient learning bounds and can solve a broader range of tasks. Additionally, we show that this framework is capable of zero-shot transfer of transition dynamics across tasks and demonstrate this empirically for the Taxi and Sokoban domains.


Reviews: Cooperative Holistic Scene Understanding: Unifying 3D Object, Layout, and Camera Pose Estimation

Neural Information Processing Systems

An approach for joint estimation of 3D Layout, 3D Object Detection, Camera Pose Estimation and Holistic Scene Understanding' (as defined in Song et al. (2015)) is proposed. More specifically, deep nets, functional mappings (e.g., projections from 3D to 2D points) and loss functions are combined to obtain a holistic interpretation of a scene illustrated in a single RGB image. The proposed approach is shown to outperform 3DGP (Choi et al. (2013)) and IM2CAD (Izadinia et al. (2017)) on the SUN RGB-D dataset. Review Summary: The paper is well written and presents an intuitive approach which is illustrated to work well when compared to two baselines. For some of the tasks, e.g., 3D Layout estimation, stronger baselines exist and as a reviewer/reader I can't assess how the proposed approach compares.