Lab Meeting September 7th, 2010 (Jimmy): Learning to Recognize Objects from Unseen Modalities

Title: Learning to Recognize Objects from Unseen Modalities
In ECCV2010

Authors: C. Mario Christoudias, Raquel Urtasun, Mathieu Salzmann and Trevor Darrell

Abstract
In this paper we investigate the problem of exploiting multiple sources of information for object recognition tasks when additional modalities that are not present in the labeled training set are available for inference. This scenario is common to many robotics sensing applications and is in contrast with the assumption made by existing approaches that require at least some labeled examples for each modality. To leverage the previously unseen features, we make use of the unlabeled data to learn a mapping from the existing modalities to the new ones. This allows us to predict the missing data for the labeled examples and exploit all modalities using multiple kernel learning. We demonstrate the e ectiveness of our approach on several multi-modal tasks including object recognition from multi-resolution imagery, grayscale and color images, as well as images and text. Our approach outperforms multiple kernel learning on the original modalities, as well as nearest-neighbor and bootstrapping schemes.

[pdf]

Lab Meeting September 7th, 2010 (Will(柏崴)): Efficient Computation of Robust Low-Rank Matrix Approximations in the Presence of Missing Data using the L1 Norm (CVPR2010)

Title: Efficient Computation of Robust Low-Rank Matrix Approximations in the Presence of Missing Data using the L1 Norm

Authors: Anders Eriksson and Anton van den Hengel

Abstract:

The calculation of a low-rank approximation of a matrix is a fundamental operation in many computer vision applications. The workhorse of this class of problems has long been the Singular Value Decomposition. However, in the presence of missing data and outliers this method is not applicable, and unfortunately, this is often the case in practice.

In this paper we present a method for calculating the low-rank factorization of a matrix which minimizes the L1 norm in the presence of missing data. Our approach represents a generalization the Wiberg algorithm, one of the more convincing methods for factorization under the L2 norm. By utilizing the differentiability of linear programs, we can extend the underlying ideas behind this approach to include this class of L1 problems as well. We show that the proposed algorithm can be efficiently implemented using existing optimization software. We also provide preliminary experiments on synthetic as well as real world data with very convincing results.

[link]

[local link]

Lab Meeting August 31st, 2010 (zhi-zhong(執中)): Efficient Planning under Uncertainty for a Target-Tracking Micro-Aerial Vehicle (ICRA'10)

Title: Efficient Planning under Uncertainty for a Target-Tracking Micro-Aerial Vehicle

Authors: Ruijie He, Abraham Bachrach and Nicholas Roy

Abstract:

A helicopter agent has to plan trajectories to track multiple ground targets from the air. The agent has partial information of each target’s pose, and must reason about its uncertainty of the targets’ poses when planning subsequent actions.

We present an online, forward-search algorithm for planning under uncertainty by representing the agent’s belief of each target’s pose as a multi-modal Gaussian belief. We exploit this parametric belief representation to directly compute the distribution of posterior beliefs after actions are taken. This analytic computation not only enables us to plan in problems with continuous observation spaces, but also allows the agent to search deeper by considering policies composed of multistep action sequences; deeper searches better enable the agent to keep the targets well-localized. We present experimental results in simulation, as well as demonstrate the algorithm on an actual quadrotor helicopter tracking multiple vehicles on a road network constructed indoors.

local copy : [link]

[link]

Lab Meeting August 31st, 2010 (David): Scene Understanding in a Large Dynamic Environment through a Laser-based Sensing (ICRA'10)

Title:

Scene Understanding in a Large Dynamic Environment through a Laser-based Sensing

Authors:

Huijing Zhao, Yiming Liu, Xiaolong Zhu, Yipu Zhao, Hongbin Zha

Abstract:

It became a well known technology that a map of complex environment containing low-level geometric primitives (such as laser points) can be generated using a robot with laser scanners. This research is motivated by the need of obtaining semantic knowledge of a large urban outdoor environment after the robot explores and generates a low-level sensing data set. An algorithm is developed with the data represented in a range image, while each pixel can be converted into a 3D coordinate. Using an existing segmentation method that models only geometric homogeneities, the data of a single object of complex geometry, such as people, cars, trees etc., is partitioned into different segments. Such a segmentation result will greatly restrict the capability of object recognition. This research proposes a framework of simultaneous segmentation and classification of range image, where the classification of each segment is conducted based on its geometric properties, and homogeneity of each segment is evaluated conditioned on each object class. Experiments are presented using the data of a large dynamic urban outdoor environment, and performance of the algorithm is evaluated.

local copy : [link]

[link]

Lab Meeting August 23rd, 2010 (Nicole): Evaluating Real-time Audio Localization Algorithms for Artificial Audition in Robotics (IROS'09)

Title: Evaluating Real-time Audio Localization Algorithms for Artificial Audition in Robotics

Authors: Anthony Badali,Jean-Marc Valin,Francois Michaud,and Parham Aarabi

Abstract:
Although research on localization of sound sources using microphone arrays has been carried out for years, providing such capabilities on robots is rather new. Artificial audition systems on robots currently exist, but no evaluation of the methods used to localize sound sources has yet been conducted. This paper presents an evaluation of various real-time audio localization algorithms using a medium-sized micro-phone array which is suitable for applications in robotics. Thetechniques studied here are implementations and enhancements of steered response power - phase transform beamformers, which represent the most popular methods for time difference of arrival audio localization. In addition, two different grid topologies for implementing source direction search are also compared. Results show that a direction refinement procedure can be used to improve localization accuracy and that more efficient and accurate direction searches can be performed using a uniform triangular element grid rather than the typical rectangular element grid.

local copy : [link]
[link]

Lab Meeting August 23rd, 2010 (ShaoChen): Distributed Nonlinear Estimation for Robot Localization using Weighted Consensus (ICRA'10)

Title: Distributed Nonlinear Estimation for Robot Localization using Weighted Consensus

Authors: Andrea Simonetto, Tam´as Keviczky and Robert Babuˇska

Abstract:

 Distributed linear estimation theory has received increased
attention  in  recent  years  due  to  several  promising
industrial applications. Distributed nonlinear estimation, however
is  still  a  relatively  unexplored  field  despite  the  need  in
numerous practical situations for techniques that can handle
nonlinearities. This paper presents a unified way of describing
distributed implementations of three commonly used nonlinear
estimators: the Extended Kalman Filter, the Unscented Kalman
Filter  and  the  Particle  Filter.  Leveraging  on  the  presented
framework,  we  propose  new  distributed  versions  of  these
methods, in which the nonlinearities are locally managed by
the various sensors whereas the different estimates are merged
based on a weighted average consensus process. The proposed
versions are shown to outperform the few published ones in
two robot localization test cases.

[link]

Lab Meeting August 10th, 2010 (KuoHuel): An Online Approach: Learning-Semantic-Scene-by-Tracking and Tracking-by-Learning-Semantic-Scene (CVPR'10)

Title: An Online Approach: Learning-Semantic-Scene-by-Tracking and

Tracking-by-Learning-Semantic-Scene

Authors: Xuan Song, Xiaowei Shao, Huijing Zhao, Jinshi Cui, Ryosuke Shibasaki and Hongbin Zha

Abstract:

Learning the knowledge of scene structure and tracking

a large number of targets are both active topics of computer

vision in recent years, which plays a crucial role in surveil-

lance, activity analysis, object classification and etc. In

this paper, we propose a novel system which simultaneously

performs the Learning-Semantic-Scene and Tracking, and

makes them supplement each other in one framework. The

trajectories obtained by the tracking are utilized to continu-

ally learn and update the scene knowledge via an online un-

supervised learning. On the other hand, the learned knowl-

edge of scene in turn is utilized to supervise and improve

the tracking results. Therefore, this “adaptive learning-

tracking loop” can not only perform the robust tracking in

high density crowd scene, dynamically update the knowl-

edge of scene structure and output semantic words, but also

ensures that the entire process is completely automatic and

online. We successfully applied the proposed system into the

JR subway station of Tokyo, which can dynamically obtain

the semantic scene structure and robustly track more than

150 targets at the same time.

[pdf]

Lab Meeting August 10th, 2010 (Jeff): FAB-MAP + RatSLAM: Appearance-based SLAM for Multiple Times of Day

Title: FAB-MAP + RatSLAM: Appearance-based SLAM for Multiple Times of Day

Authors: Arren J. Glover, William P. Maddern, Michael J. Milford, and Gordon F. Wyeth

Abstract:

Appearance-based mapping and localisation is especially challenging when separate processes of mapping and localisation occur at different times of day. The problem is exacerbated in the outdoors where continuous change in sun angle can drastically affect the appearance of a scene. We confront this challenge by fusing the probabilistic local feature based data association method of FAB-MAP with the pose cell filtering and experience mapping of RatSLAM. We evaluate the effectiveness of our amalgamation of methods using five datasets captured throughout the day from a single camera driven through a network of suburban streets. We show further results when the streets are re-visited three weeks later, and draw conclusions on the value of the system for lifelong mapping.

Link:
IEEE International Conference on Robotics and Automation(ICRA), May 2010
http://eprints.qut.edu.au/31569/1/c31569.pdf
or
local_copy

CVPR 2010 Awards

This post is to provide links to the best paper awards in CVPR 2010.

Best Student Paper

• Visual Event Recognition in Videos by Learning from Web Data: Lixin Duan, Dong Xu, Ivor Wai-Hung Tsang, and Jiebo Luo

Best Paper Honorable Mention

• Modeling Mutual Context of Object and Human Pose in Human-Object Interaction Activities: Bangpeng Yao and Li Fei-Fei

Best Paper

• Efficient Computation of Robust Low-Rank Matrix Approximations in the Presence of Missing Data using the L1 Norm: Anders Eriksson and Anton van den Hengel

Longuet-Higgins Prize

• Efficient Matching of Pictorial Structures: Pedro F. Felzenszwalb and Daniel P. Huttenlocher
• Real-Time Tracking of Non-Rigid Objects Using Mean Shift: Dorin Comaniciu, Visvanathan Ramesh, and Peter Meer

Lab Meeting August 3rd, 2010 (Wang Li): Modeling Mutual Context of Object and Human Pose in Human-Object Interaction Activities (CVPR 2010)

Modeling Mutual Context of Object and Human Pose in Human-Object Interaction Activities

Bangpeng Yao
Li Fei-Fei

Abstract
Detecting objects in cluttered scenes and estimating articulated human body parts are two challenging problems in computer vision. We observe, however, that objects and human poses can serve as mutual context to each other – recognizing one facilitates the recognition of the other.
In this paper, we propose a new random field model to encode the mutual context of objects and human poses in human-object interaction activities. We then cast the model learning task as a structure learning problem, of which the structural connectivity between the object, the overall human pose and different body parts are estimated through a structure search approach, and the parameters of the model are estimated by a new max-margin algorithm.
On a sports data set of six classes of human-object interactions, we show that our mutual context model significantly outperforms state-of-the-art in detecting very difficult objects and human poses.

Paper Link

Lab Meeting 8 / 3, 2010 (Alan) - Mapping Indoor Environments Based on Human Activity (ICRA 2010)

Title: Mapping Indoor Environments Based on Human Activity (ICRA 2010)

Authors: Slawomir Grzonka, Frederic Dijoux, Andreas Karwath, Wolfram Burgard

Abstract

We present a novel approach to build approximate maps of structured environments utilizing human motion and activity. Our approach uses data recorded with a data suit which is equipped with several IMUs to detect movements of a person and door opening and closing events. In our approach we interpret the movements as motion constraints and door handling events as landmark detections in a graph-based SLAM framework. As we cannot distinguish between individual doors, we employ a multi-hypothesis approach on top of the SLAM system to deal with the high data-association uncertainty. As a result, our approach is able to accurately and robustly recover the trajectory of the person. We additionally take advantage of the fact that people traverse free space and that doors separate rooms to recover the geometric structure of the environment after the graph optimization. We evaluate our approach in several experiments carried out with different users and in environments of different types.

Link: pdf

Lab Meeting 07/27, 2010(Kuen-Han) Non-Rigid Structure from Locally-Rigid Motion (CVPR,2010)

Title: Non-Rigid Structure from Locally-Rigid Motion
Authors: Jonathan Taylor Allan D. Jepson Kiriakos N. Kutulakos

Abstract:
We introduce locally-rigid motion, a general framework for
solving the M-point, N-view structure-from-motion problem
for unknown bodies deforming under orthography. The
key idea is to first solve many local 3-point, N-view rigid
problems independently, providing a “soup” of specific,
plausibly rigid, 3D triangles. The main advantage here is
that the extraction of 3D triangles requires only very weak
assumptions: (1) deformations can be locally approximated
by near-rigid motion of three points (i.e., stretching not
dominant) and (2) local motions involve some generic rotation
in depth. Triangles from this soup are then grouped
into bodies, and their depth flips and instantaneous relative
depths are determined. Results on several sequences,
both our own and from related work, suggest these conditions
apply in diverse settings—including very challenging
ones (e.g., multiple deforming bodies). Our starting point
is a novel linear solution to 3-point structure from motion,
a problem for which no general algorithms currently exist.

paper

Lab Meeting July 20, 2010 (fish60): What if the Irresponsible Teachers Are Dominating? A Method of Training on Samples and Clustering on Teachers

Sorry for the previous blank post.
Here's the content:

Title:
What if the Irresponsible Teachers Are Dominating? A Method of Training on Samples and Clustering on Teachers

Authors:
Shuo Chen, Jianwen Zhang, Guangyun Chen, Changshui Zhang
State Key Laboratory on Intelligent Technology and Systems
Tsinghua National Laboratory for Information Science and Technology (TNList)
Department of Automation, Tsinghua University, Beijing 100084, China

Abstract:
Learning from multiple teachers or sources
has received more attention of the researchers in the machine
learning area. In this setting, the learning system is dealing
with samples and labels provided by multiple teachers, who
in common cases, are non-expert. Their labeling styles and
behaviors are usually diverse, some of which are even detrimental
to the learning system. Thus, simply putting them
together and utilizing the algorithms designed for singleteacher
scenario would be not only improper, but also damaging.
Our work focuses on a case where the teachers are composed of good
ones and irresponsible ones. By irresponsible, we mean the
teacher who takes the labeling task not seriously and label
the sample at random without inspecting the sample itself.
If we do not take out their effects, our learning system would be ruined with no
doubt. In this paper, we propose a method for picking out the
good teachers with promising experimental results. It works
even when the irresponsible teachers are dominating in numbers.

Link

江山代有才人出 攻讀博士—不輕言放棄

Author: 王榮騰 臺大客座教授

即使興趣不完全符合指導教授之研究領域，應考慮與其適時溝通、更換原指定之研究項目，甚至要求同時加入另一教授研究團隊，仍繼續跟定指導教授，不輕言放棄，或才是良策。

A生曾榮獲美國極頂尖大學某一指導教授〈Advisor〉給予的全額研究助理獎學金〈RA，Research Assistantship〉，一年半後，A生放棄學業，正在覓職中！
B生曾榮獲美國另一極頂尖大學給予的一年期全額研究生獎學金〈Graduate Fellowship〉，一年後及時拿到RA，卻一直認為研究與現實脫節，擔心未來就業機會而深感困擾！

這兩位高材生皆因成績優異，才能獲美國極頂尖大學給予的全額獎學金，卻皆因興趣不全在指導教授研究領域，而心蒙去念或煩惱不已！

其實，人生不如意十有八九，人生本就不完美，古今中外皆然。即使未來就職、自己創業或擔任教授，又怎能保證事事如意。已是天之驕子，若不能克服眼前困難，輕言放棄，今後何以立足於社會？因此，提供一些建議，學習如何解決問題，創造師生雙贏局面。



續跟指導教授
------------更換研究項目

指導教授常會同時進行數個研究項目，可當面請教並說明原因，是否能更換原指定之研究題目。若非不合理，教授多半都會接受。須知，博士論文〈PhD Dissertation〉大多由幾個研究專題組合而成。因此，最好是在文章被期刊或會議接受後提出；一來，可對目前該專題有所交代〈不至於浪費教授研究經費〉，二來也有助於自己博士論文的進展。再者，亦可利用這段時間對新研究項目有所了解。

假如研究題目太難以致研究上難已突破並發表文章時，更須謹言並舉例佐證，以免教授懷疑你的能力，對你失去信心。如純粹是興趣不足，最好是在老師指定題目數週內及時告知，以免留下不好印象。無論如何，必須同時向教授提出你的興趣所在，並告知你的背景及觀點，以便讓他對你能在此新研究專題上有所突破更具信心。



另加指導教授
-------------研究跨領域項目

對跨領域的新研究專題，可請求老師，是否能同時加入另一共同指導教授〈Co-Advisor〉的研究團隊。這項權宜之計，由於兩位教授可平均分擔研究經費，卻同時受惠於你未來的研究成果，因此指導教授通常會樂於接受。

換言之，不管個人興趣是否與指導教授研究領域相近，繼續跟定指導教授，不輕言放棄；且莫在博士資格考試〈PhD Qualifying Examination〉未通過前提出，以免造成輟學的嚴重後果。

經過長期溝通，最後A生接受指導教授建議，先休學、工作一段時間，再考慮是否繼續完成博士學位。B生則同時加入另一教授之研究團隊，不排除於畢業後往學術界發展；其後續已不再為研究課題而煩惱，並已在新覓研究領域之尖端會議中發表論文。由於處理得宜，目前這兩位高材生仍與原來指導教授保持良好關係。畢竟，恩師難覓，必須知福惜福；師生情難建，值得一生珍惜！ 〈王榮騰 臺大電機系與電子工程研究所客座教授;2010年6月6日〉

Lab Meeting July 20, 2010 (Gary): Robust Unified Stereo-Based 3D Head Tracking and Its Application to Face Recognition (ICRA2010)

Title:
Robust Unified Stereo-Based 3D Head Tracking and Its Application
to Face Recognition

Authors: Kwang Ho An and Myung Jin Chung

Abstract:
This paper investigates the estimation of 3D head poses and its identity authentication with a partial ellipsoid model. To cope with large out-of-plane rotations and translation in-depth, we extend conventional head tracking with a single camera to a stereo-based framework. To achieve more robust motion estimation even under time-varying lighting conditions, we incorporate illumination correction into the aforementioned framework. We approximate the face image variations due to illumination changes as a linear combination of illumination bases. Also,��by computing the illumination bases online from the registered face images, after estimating the 3D head poses, user-specific illumination bases can be obtained, and therefore illumination-robust tracking without a prior learning process can be possible. Furthermore, our unified stereo-based tracking is approximated as a linear least-squares problem; a closed-form solution is then provided. After recovering the full-motions of the head, we can register face images with pose variations into stabilized-view images, which are suitable for pose-robust face recognition. To verify the feasibility and applicability of our approach, we performed extensive experiments with three sets of challenging image sequences.

link

Lab Meeting July 20, 2010 (Jimmy): Group-Sensitive Multiple Kernel Learning for Object Categorization

Title: Group-Sensitive Multiple Kernel Learning for Object Categorization
Authors: Jingjing Yang, Yuanning Li, Yonghong Tian, Lingyu Duan, Wen Gao
In: ICCV 2009

Abstract
In this paper, we propose a group-sensitive multiple kernel learning (GS-MKL) method to accommodate the intra-class diversity and the inter-class correlation for object categorization. By introducing an intermediate representation “group” between images and object categories, GS-MKL attempts to find appropriate kernel combination for each group to get a finer depiction of object categories. For each category, images within a group share a set of kernel weights while images from different groups may employ distinct sets of kernel weights. In GS-MKL, such group-sensitive kernel combinations together with the multi-kernels based classifier are optimized in a joint manner to seek a trade-off between capturing the diversity and keeping the invariance for each category. Extensive experiments show that our proposed GS-MKL method has achieved encouraging performance over three challenging datasets.

[pdf]

Lab Meeting July 13, 2010(ShaoChen):Rao-Blackwellized Particle Filters Multi Robot SLAM with Unknown Initial Correspondences and Limited Communication(ICRA 2010)

Title: Rao-Blackwellized Particle Filters Multi Robot SLAM with Unknown Initial Correspondences and Limited Communication

Authors: Luca Carlone, Miguel Kaouk Ng, Jingjing Du, Basilio Bona, and Marina Indri

Abstract:

Multi robot systems are envisioned to play an important role in many robotic applications. A main prerequisite for a team deployed in a wide unknown area is the capability of autonomously navigate, exploiting the information acquired through the on-line estimation of both robot poses
and surrounding environment model, according to Simultaneous Localization And Mapping (SLAM) framework. As team coordination is improved, distributed techniques for filtering
are required in order to enhance autonomous exploration and large scale SLAM increasing both efficiency and robustness of operation. Although Rao-Blackwellized Particle Filters (RBPF) have been demonstrated to be an effective solution to the problem of single robot SLAM, few extensions to teams of robots exist, and these approaches are characterized by strict assumptions on both communication bandwidth and prior knowledge on relative poses of the teammates. In the present paper we address the problem of multi robot SLAM in the case of limited communication and unknown relative initial poses. Starting from the well established single robot RBPFSLAM, we propose a simple technique which jointly estimates SLAM posterior of the robots by fusing the prioceptive and the eteroceptive information acquired by each teammate. The approach intrinsically reduces the amount of data to be exchanged among the robots, while taking into account the uncertainty in relative pose measurements. Moreover it can be naturally extended to different communication technologies (bluetooth, RFId, wifi, etc.) regardless their sensing range. The proposed approach is validated through experimental test.

[link]

Lab Meeting July 13,2010(Nicole):Mutual Localization in a Team of Autonomous Robots using Acoustic Robot Detection

Title: Mutual Localization in a Team of Autonomous Robots using Acoustic Robot Detection

Authors: David Becker and Max Risler

In RoboCup 2008: Robot Soccer World Cup XII ,Volume 5399/2009


Abstract
In order to improve self-localization accuracy we are exploring ways of mutual localization in a team of autonomous robots. Detecting team mates visually usually leads to inaccurate bearings and only rough distance estimates. Also, visually identifying teammates is not possible. Therefore we are investigating methods of gaining relative position information acoustically in a team of robots.
The technique introduced in this paper is a variant of code-multiplexed communication (CDMA, code division multiple access). In a CDMA system, several receivers and senders can communicate at the same time, using the same carrier frequency. Well-known examples of CDMA systems include wireless computer networks and the Global Positioning System, GPS. While these systems use electro-magnetic waves, we will try to adopt the CDMA principle towards using acoustic pattern recognition, enabling robots to calculate distances and bearings to each other.
First, we explain the general idea of cross-correlation functions and appropriate signal pattern generation. We will further explain the importance of synchronized clocks and discuss the problems arising from clock drifts.
Finally, we describe an implementation using the Aibo ERS-7 as platform and briefly state basic results, including measurement accuracy and a runtime estimate. We will briefly discuss acoustic localization in the specific scenario of a RoboCup soccer game.

[link]

Lab Meeting July 6th (Casey): Live Dense Reconstruction with a Single Moving Camera (CVPR 2010)

Authors: Richard A. Newcombe and Andrew J. Davison

Abstract:

We present a method which enables rapid and dense reconstruction of scenes browsed by a single live camera. We take point-based real-time structure from motion (SFM) as our starting point, generating accurate 3D camera pose estimates and a sparse point cloud. Our main novel contribution is to use an approximate but smooth base mesh generated from the SFM to predict the view at a bundle of poses around automatically selected reference frames spanning the scene, and then warp the base mesh into highly accurate depth maps based on view-predictive optical flow and a constrained scene flow update. The quality of the resulting depth maps means that a convincing global scene model can be obtained simply by placing them side by side and removing overlapping regions. We show that a cluttered indoor environment can be reconstructed from a live hand-held camera in a few seconds, with all processing performed by current desktop hardware. Real-time monocular dense reconstruction opens up many application areas, and we demonstrate both real-time novel view synthesis and advanced augmented reality where augmentations interact physically with the 3D scene and are correctly clipped by occlusions.

[Paper]

Lab Meeting July 6th 2010 (Andi): Upsampling Range Data in Dynamic Environments (CVPR 2010 )

Authors

Jennifer Dolson, Jongmin Baek, Christian Plagemann and Sebastian Thrun (Stanford University)

Abstract

We present a flexible method for fusing information from optical and range sensors based on an accelerated high-dimensional filtering approach. Our system takes as input a sequence of monocular camera images as well as a stream of sparse range measurements as obtained from a laser or other sensor system. In contrast with existing approaches, we do not assume that the depth and color data streams have the same data rates or that the observed scene is fully static. Our method produces a dense, high-resolution depth map of the scene, automatically generating confidence values for every interpolated depth point. We describe how to integrate priors on object shape, motion and appearance and how to achieve an efficient implementation using parallel processing hardware such as GPUs.

paper