丕子

研究：Twitter情感趋势反映Facebook IPO走势

丕子 — Mon, 21 May 2012 07:38:50 +0800

本文转载自：36氪

股价的走势历来都让人琢磨不透。但是Facebook IPO当天，Twitter却神奇般的预测到了其股价变动的每一个趋势。

社交媒体监测平台DataSift监测了Facebook IPO当天Twitter上的情感倾向。即人们在Twitter上发布的大量谈论facebook IPO的Tweet中会包含一些情感倾向，比如正面的或者负面的。结论显示，Twitter上每一次情感倾向的转向都会影响Facebook股价的波动，延迟情况只有几分钟到二十多分钟。

DataSift根据58665位用户产生的95019条Tweet互动绘制了一幅图，很清晰的反映了两者的对应关系——基本上是呈正相关。

比如，Facebook 开盘前Twitter上的情感逐渐转向负面，25分钟之后,Facebook的股价便开始下跌。而当Twitter上的情感转向正面时，Facebook股价在8分钟之后也开始了回弹。最后，当股市接近收盘时，Twitter上的情感转向负面，10分钟后facebook的股价又开始下跌。

当然，就凭DataSift的这幅图还远不能揭示Twitter和股价的关系，也不能证明Twitter会对股价产生影响。不过国外和国内确实已经开始了对社交媒体进行情感语义分析，并试图由此发现另外一条预测股价趋势的方法。对于社交媒体和股票感兴趣的创业者可以研究一下。

毕业临近，旅旅游

丕子 — Sun, 20 May 2012 00:01:44 +0800

最近一段时间，毕业在即，各种琐事，明天毕业答辩。

最可恶的就是驾校由于人太多，考试进度非常非常慢，耽误了很多事情，也没办法出去实习。等吧。

由于有点儿疲，就索性背起包来连夜爬了个泰山，之前种种原因错过了两次爬泰山的机会。

运气不错，看到了日观峰的日出。

宿舍两个兄弟都已经有驾照了，索性租了一辆小车，自驾游吧，去了三个同学的家。

兄弟们的父母都很热情，很亲近，自己也有到家的一种感觉，反而没有那么陌生。

去了临沂，爬了蒙山，和泰山真是各有风格。

去了枣庄滕州，泛舟微山湖，真是大，一望无际，到了京杭大运河的分支漂了一下，环境真是好。

近距离接触了渔民的生活。

吃了家里各种地道的饭菜，胖了。

===

6月4号奔赴香港开一个小会，除了开会，更想看看香港的特色，自己以后开会的机会也少了，也没写出什么像样子的文章。

自己常常反思，研究生阶段过的有点儿浮躁，有点儿虚无，不是很踏实。

===

毕业，等待入职后新的开始。

自己进入状态慢，需要加油了。

PMTK3 :probabilistic modeling toolkit for Matlab/Octave

丕子 — Mon, 07 May 2012 07:53:53 +0800

是书《Machine Learning: a Probabilistic Perspective》的代码：

A new textbook by Kevin Murphy (in press, should be available by the end of August 2012)

My book (MLaPP) is similar to Bishop's Pattern recognition and machine learning, Hastie et al's The Elements of Statistical Learning, and to Wasserman's All of statistics, with the following key differences:

A list of every PMTK3 demo

Updated Mar 27, 2012 by mattbdun...@gmail.com

auto-generated by publishDemos on 27-Mar-2012

Book Demos

Other Demos

a collection of demos not appearing in the book

OtherDemos

一个库，一个工具箱

丕子 — Fri, 04 May 2012 11:42:22 +0800

from：http://people.csail.mit.edu/dhlin/softwares.html

Basic Computation Supporting Library (BCSLib)

This is a C++ library to provide basic support to various computation tasks. The library comprises the following modules:

Base module: basic algorithms, memory management support, and meta-programming support
Array module: template classes to represent vectors, matrices, vector and matrix views using different indexing, as well as functions to manipulate vectors and matrices.
Data structure module: template classes and functions that implement various data structures, such as heaps, disjoint sets, and different types of trees.
Graph module: template classes to represent graphs, and generic graph algorithms including BFS, DFS, topological sorting, shortest path algorithms, and minimum spanning tree algorithms.

Code website: http://code.google.com/p/bcslib/

Statistical Modeling and Inference Toolbox

This is a MATLAB Toolbox to support statistical learning and inference. It comprises over 300 functions, ranging from classic algorithms such as Principle Component Analysis (PCA) and K-Means, to advanced probabilistic models such as Latent Dirichlet Allocation and nonparametric Bayesian. Specifically, the toolbox is composed of the following modules:

Base module: basic but useful functions for data manipulation, pairwise metric computation, and efficient routines to deal with the calculation on large amount of small matrices.
Optim module: functions for nonlinear optimization (e.g. Newton and BFGS), and wrappers for LP\QP solvers, as well as functions to solve optimization problems of specific forms
Classic module: a collection of classic algorithms
- KNN-classification, K-means and K-medoid clustering, ROC Analysis
- Linear regression, Logistic regression, Robust regression, Generic regularized loss minimization
- Subspace learning: Principal Component Analysis, Fisher Linear Discriminant Analysis, Independent Component Analysis
- Support vector machines for classification, regression, and ranking.
Graph module: a collection of functions to solve graph-related problems.
- Functions to efficiently represent and manipulate graphs
- Classic graph algorithms: BFS, DFS, Connect components, Shortest paths, Minimum spanning trees, etc.
- Network flow algorithms: maximum flow minimum cut.
- Algorihms based on spectral graph theory: Graph embedding and normalized cut.
P-Model module: classes and functions for probabilistic modeling and inference.
- Evaluation and sampling of well-known distributions: Gaussian distribution, Gamma distribution, Dirichlet distribution, Wishart distribution, and inverse Wishart distribution, etc. This toolbox provides more efficient implementation than MATLAB's statistics toolbox on many functions.
- Markov models: Markov chains, Markov random fields and Hidden Markov Models
- Finite mixture models and Dirichlet process Mixture models.
- Topic models, such as Latent Dirichlet Allocation.

Code website: http://code.google.com/p/smitoolbox/

人生苍茫，最近状态

丕子 — Sun, 29 Apr 2012 21:22:48 +0800

真是发现不学习不思考就没得写了，只能写写生活。

最近没啥事情，主要的工作就是学车考驾照，由于驾校人太多所以考试周期跨度太大了，4月14考了科目二，科目三得5月考。要是不考驾照也就去实习了吧。

要穷死了，由于今年没啥大额的奖学金啥的，也没出去实习，也没私活，然后就囧了，再者要是去北京工作又得押一付三，看来得向哥们儿求助了。

一穷二白，一屁股债，北上打工，谁知道命途如何，我又不是个聪明人。

忽然发现好多好多同学去银行了，大多是为了生活，生活也挺好的。

为了坚持自己的一点儿小梦想，银行、公务员啥的也没考虑过，希望以后不会后悔。

其实选择没有错，好好做呗，到死了一直做得挺好的话也值了，本来就不可能体验所有的职业和生活。

或许是人设计的一种优化，又或者是一种Bug。

同志们，假期过得愉快哈，多陪陪父母妻儿，生命短暂。

Accepted Papers of ICMR 2012

丕子 — Thu, 19 Apr 2012 15:25:02 +0800

ICMR: ACM International Conference on Multimedia Retrieval。这个会议是以前的ACM CIVR 和ACM MIR合并而成，关注图像、音频、视频的检索。

会议文章比较少，慢慢发展吧。

Tentative schedule:

5th, June       : Tutorials
6~8th, June   : Main conference program
8th, June       : Practitioner day

Accepted Papers:

ID	Title	Paper Type
Regular paper
2	Visual Pattern Discovery for Architecture Image Classification and Product Image Search	Poster
3	Learning to Summarize Web Image and Text Mutually	Poster
7	Classification based Group Photo Retrieval with Bag of People Features	Oral
8	An Extensible Personal Photograph Collection for Graded Relevance Assessments and User Simulation	Poster
10	MUSICAL KEYS AND CHORDS DETECTION USING SEMI-SUPERVISED LEARNING WITH INFINITE GAUSSIAN MIXTURE	Poster
13	Contour Canonical Form: An Efficient Intrinsic Embedding Approach to Matching Non-rigid 3D Objects	Poster
17	Semiconducting Bilinear Deep Learning for Incomplete Image Recognition	Poster
19	Color CENTRIST: A Color Descriptor for Scene Categorization	Poster
23	Name that sculpture	Oral
24	Video Retrieval by Mimicking Poses	Poster
26	Ordinal Preserving Projection: A Novel Dimensionality Reduction Method for Image Ranking	Oral
30	Multi-Graph Multi-Instance Learning for Object-Based Image and Video Retrieval	Poster
32	High-Confidence Near-Duplicate Image Detection	Oral
38	Content is Still King: The Effect of Neighbor Voting Schemes on Tag Relevance for Social Image Retrieval	Oral
40	BECAUSE NOT ALL DISPLAYS ARE LISTS!	Oral
41	Efficient Graffiti Image Retrieval	Poster
46	Joint-Rerank: A Novel Method for Image Search Reranking	Poster
55	Multimodal Feature Generation Framework for Semantic Image Classification	Poster
61	Joint Audio-Visual Bi-Modal Codewords for Video Event Detection	Poster
63	SUPER: Towards Real-time Event Recognition in Internet Videos	Oral
64	ImageTerrier: An extensible platform for scalable high-performance image retrieval	Poster
67	Indexing Media by Personal Events	Poster
68	Labelset Anchored Subspace Ensemble (LASE) for Multi-label Annotation	Poster
75	Security-Oriented Picture-In-Picture Visual Modifications	Oral
76	Exploring Two Spaces With One Feature: Kernelized Multidimensional Modeling of Visual Alphabets	Oral
80	Efficient Video Copy Detection via Aligning Video Signature Time Series	Oral
86	Fusing Concept Detection and Geo Context for Visual Search	Oral
91	Constrained Keypoint Quantization: Towards Better Bag-of-Words Model for Large-scale Multimedia Retrieval	Oral
92	Bayesian approach for near-duplicate image detection	Oral
94	Distributed KNN-Graph approximation via Hashing	Poster
99	Analysing Facebook Features to Support Event Detection for Photo-Based Facebook Applications	Oral
114	Deriving Discriminative Color Model for a Given Object Class from Weakly Labeled Training Data	Poster
115	Quantity versus Quality – The role of layout complexity in thumbnail based video retrieval interfaces	Poster
123	Multi-Modal Region Selection Approach for Training Object Detectors	Oral
132	Geo-based Automatic Image Annotation	Poster
133	Compact Hashing for Mixed Image-Keyword Query over Multi-Label Images	Oral
135	Watching and Talking: Media Content as Social Nexus	Oral
136	Boosting Saliency Detection with Robust Temporal Alignment and Local-Global Spatial Contrast	Poster
138	Discovering Inherent Event Taxonomies from Social Media Collections	Poster
139	Advanced shape context for plant species identification using leaf image retrieval	Poster
149	Classifier-specific Intermediate Representation for Multimedia Tasks	Poster
152	Multimodal Knowledge-based Analysis in Multimedia Event Detection	Poster
154	Large Vocabulary Quantization for Searching Instances from Videos	Poster
155	Beyond Audio and Video Retrieval: Towards Multimedia Summarization	Oral
156	Hamming Embedding Similarity-based Image Classification	Oral
157	A Visual Approach for Video Geocoding using Bag-of-Scenes	Poster
160	A RELIEF-Based Modality Weighting Approach for Multimodal Information Retrieval	Poster
161	Supervised Dictionary Learning for Music Information Retrieval	Poster
168	All Vehicles are Cars: Subclass Preferences in Container Concepts	Oral
170	Image Retrieval Using Multimodal Fusion based on Matrix Factorization	Poster
Special Session 1: Social Events in Web Multimedia (oral session)
22	Social Event Detection and Retrieval in Collaborative Photo Collections	Oral
73	Event-based Classification of Social Media Streams	Oral
113	Social Event Detection using Multimodal Clustering and Integrating Supervisory Signals	Oral
Special Session 2: Socio-Video Semantics (oral session)
56	Linking Visual Concept Detection with Viewer Demographics	Oral
107	Cross-Modal Categorisation of User-Generated Video Sequences	Oral
147	Making a Scene: Alignment of Complete Sets of Clips based on Pairwise Audio Match	Oral
Demos
9	nepDroid: An Intelligent Mobile Music Player	Demo
21	One Size Does Not Fit All – Multimodal Search on Mobile and Desktop Devices with the I-SEARCH Search Engine	Demo
51	PythiaSearch -- A Multiple Search Strategy-supportive Multimedia Retrieval System	Demo
72	Cluster-based Photo Browsing and Tagging on the Go	Demo
78	Mobile Image Browsing on a 3D Globe	Demo
87	Categorization of a collection of pictures into structured events	Demo
90	Texmix: An automatically generated news navigation portal	Demo
122	Supporting Browsing of User Generated Video on a Tablet	Demo
128	World Seer : A Realtime Geo-Tweet Photo Mapping System	Demo
134	Image Exploration Using Online Feature Extraction and Reranking	Demo
169	Active learning of custom sound taxonomies in unstructured audio data	Demo

最近单曲循环--《再见青春》

丕子 — Mon, 16 Apr 2012 08:27:30 +0800

越听越有感觉，越听心情越低落。。。

我将在深秋的黎明出发
伴着铁皮车厢的摇晃
伴着野菊花开的芬芳
在梦碎的黎明出发

再见青春
再见美丽的疼痛
再见青春
永恒的迷惘

雨会从记忆的指间滑落
带着血中曼舞的青鸟
带着风中悲鸣的草帽
从燃烧的风中滑落

再见青春
再见美丽的疼痛
再见青春
永远的故乡

再见青春
再见灿烂的忧伤
再见青春
永恒的迷惘

我曾随迷失的航船沉没
陷入璀璨虚空的碎梦
沉入乱欲冰封的深谷
随烂漫的星群沉没

我看着满目创痍的繁华
感到痛彻心肺的惆怅
听着心在爆裂的巨响
陷入深不见底的悲伤

再见青春
再见美丽的疼痛
再见青春
永远的故乡

再见青春
再见灿烂的忧伤
再见青春
永恒的迷惘

最热文章

收藏：Topic modeling made just simple enough

丕子 — Sun, 15 Apr 2012 20:10:17 +0800

收藏文章来自于：http://www.52nlp.cn。

原文见：Topic modeling made just simple enough

Topic modeling made just simple enough.

Posted on April 7, 2012

Right now, humanists often have to take topic modeling on faith. There are several good posts out there that introduce the principle of the thing (by Matt Jockers, for instance, and Scott Weingart). But it’s a long step up from those posts to the computer-science articles that explain “Latent Dirichlet Allocation” mathematically. My goal in this post is to provide a bridge between those two levels of difficulty.

Computer scientists make LDA seem complicated because they care about proving that their algorithms work. And the proof is indeed brain-squashingly hard. But the practice of topic modeling makes good sense on its own, without proof, and does not require you to spend even a second thinking about “Dirichlet distributions.” When the math is approached in a practical way, I think humanists will find it easy, intuitive, and empowering. This post focuses on LDA as shorthand for a broader family of “probabilistic” techniques. I’m going to ask how they work, what they’re for, and what their limits are.

How does it work? Say we’ve got a collection of documents, and we want to identify underlying “topics” that organize the collection. Assume that each document contains a mixture of different topics. Let’s also assume that a “topic” can be understood as a collection of words that have different probabilities of appearance in passages discussing the topic. One topic might contain many occurrences of “organize,” “committee,” “direct,” and “lead.” Another might contain a lot of “mercury” and “arsenic,” with a few occurrences of “lead.” (Most of the occurrences of “lead” in this second topic, incidentally, are nouns instead of verbs; part of the value of LDA will be that it implicitly sorts out the different contexts/meanings of a written symbol.)

Of course, we can’t directly observe topics; in reality all we have are documents. Topic modeling is a way of extrapolating backward from a collection of documents to infer the discourses (“topics”) that could have generated them. (The notion that documents are produced by discourses rather than authors is alien to common sense, but not alien to literary theory.) Unfortunately, there is no way to infer the topics exactly: there are too many unknowns. But pretend for a moment that we had the problem mostly solved. Suppose we knew which topic produced every word in the collection, except for this one word in document D. The word happens to be “lead,” which we’ll call word type W. How are we going to decide whether this occurrence of W belongs to topic Z?

We can’t know for sure. But one way to guess is to consider two questions. A) How often does “lead” appear in topic Z elsewhere? If “lead” often occurs in discussions of Z, then this instance of “lead” might belong to Z as well. But a word can be common in more than one topic. And we don’t want to assign “lead” to a topic about leadership if this document is mostly about heavy metal contamination. So we also need to consider B) How common is topic Z in the rest of this document?

Here’s what we’ll do. For each possible topic Z, we’ll multiply the frequency of this word type W in Z by the number of other words in document D that already belong to Z. The result will represent the probability that this word came from Z. Here’s the actual formula:

Simple enough. Okay, yes, there are a few Greek letters scattered in there, but they aren’t terribly important. They’re called “hyperparameters” — stop right there! I see you reaching to close that browser tab! — but you can also think of them simply as fudge factors. There’s some chance that this word belongs to topic Z even if it is nowhere else associated with Z; the fudge factors keep that possibility open. The overall emphasis on probability in this technique, of course, is why it’s called probabilistic topic modeling.

Now, suppose that instead of having the problem mostly solved, we had only a wild guess which word belonged to which topic. We could still use the strategy outlined above to improve our guess, by making it more internally consistent. We could go through the collection, word by word, and reassign each word to a topic, guided by the formula above. As we do that, a) words will gradually become more common in topics where they are already common. And also, b) topics will become more common in documents where they are already common. Thus our model will gradually become more consistent as topics focus on specific words and documents. But it can’t ever become perfectly consistent, because words and documents don’t line up in one-to-one fashion. So the tendency for topics to concentrate on particular words and documents will eventually be limited by the actual, messy distribution of words across documents.

That’s how topic modeling works in practice. You assign words to topics randomly and then just keep improving the model, to make your guess more internally consistent, until the model reaches an equilibrium that is as consistent as the collection allows.

What is it for? Topic modeling gives us a way to infer the latent structure behind a collection of documents. In principle, it could work at any scale, but I tend to think human beings are already pretty good at inferring the latent structure in (say) a single writer’s oeuvre. I suspect this technique becomes more useful as we move toward a scale that is too large to fit into human memory.

So far, most of the humanists who have explored topic modeling have been historians, and I suspect that historians and literary scholars will use this technique differently. Generally, historians have tried to assign a single label to each topic. So in mining the Richmond Daily Dispatch, Robert K. Nelson looks at a topic with words like “hundred,” “cotton,” “year,” “dollars,” and “money,” and identifies it as TRADE — plausibly enough. Then he can graph the frequency of the topic as it varies over the print run of the newspaper.

As a literary scholar, I find that I learn more from ambiguous topics than I do from straightforwardly semantic ones. When I run into a topic like “sea,” “ship,” “boat,” “shore,” “vessel,” “water,” I shrug. Yes, some books discuss sea travel more than others do. But I’m more interested in topics like this:

You can tell by looking at the list of words that this is poetry, and plotting the volumes where the topic is prominent confirms the guess.

This topic is prominent in volumes of poetry from 1815 to 1835, especially in poetry by women, including Felicia Hemans, Letitia Landon, and Caroline Norton. Lord Byron is also well represented. It’s not really a “topic,” of course, because these words aren’t linked by a single referent. Rather it’s a discourse or a kind of poetic rhetoric. In part it seems predictably Romantic (“deep bright wild eye”), but less colorful function words like “where” and “when” may reveal just as much about the rhetoric that binds this topic together.

A topic like this one is hard to interpret. But for a literary scholar, that’s a plus. I want this technique to point me toward something I don’t yet understand, and I almost never find that the results are too ambiguous to be useful. The problematic topics are the intuitive ones — the ones that are clearly about war, or seafaring, or trade. I can’t do much with those.

Now, I have to admit that there’s a bit of fine-tuning required up front, before I start getting “meaningfully ambiguous” results. In particular, a standard list of stopwords is rarely adequate. For instance, in topic-modeling fiction I find it useful to get rid of at least the most common personal pronouns, because otherwise the difference between 1st and 3rd person point-of-view becomes a dominant signal that crowds out other interesting phenomena. Personal names also need to be weeded out; otherwise you discover strong, boring connections between every book with a character named “Richard.” This sort of thing is very much a critical judgment call; it’s not a science.

I should also admit that, when you’re modeling fiction, the “author” signal can be very strong. I frequently discover topics that are dominated by a single author, and clearly reflect her unique idiom. This could be a feature or a bug, depending on your interests; I tend to view it as a bug, but I find that the author signal does diffuse more or less automatically as the collection expands.

What are the limits of probabilistic topic modeling? I spent a long time resisting the allure of LDA, because it seemed like a fragile and unnecessarily complicated technique. But I have convinced myself that it’s both effective and less complex than I thought. (Matt Jockers, Travis Brown, Neil Fraistat, and Scott Weingart also deserve credit for convincing me to try it.)

This isn’t to say that we need to use probabilistic techniques for everything we do. LDA and its relatives are valuable exploratory methods, but I’m not sure how much value they will have as evidence. For one thing, they require you to make a series of judgment calls that deeply shape the results you get (from choosing stopwords, to the number of topics produced, to the scope of the collection). The resulting model ends up being tailored in difficult-to-explain ways by a researcher’s preferences. Simpler techniques, like corpus comparison, can answer a question more transparently and persuasively, if the question is already well-formed. (In this sense, I think Ben Schmidt is right to feel that topic modeling wouldn’t be particularly useful for the kinds of comparative questions he likes to pose.)

Moreover, probabilistic techniques have an unholy thirst for memory and processing time. You have to create several different variables for every single word in the corpus. The models I’ve been running, with roughly 2,000 volumes, are getting near the edge of what can be done on an average desktop machine, and commonly take a day. To go any further with this, I’m going to have to beg for computing time. That’s not a problem for me here at Urbana-Champaign (you may recall that we invented HAL), but it will become a problem for humanists at other kinds of institutions.

Probabilistic methods are also less robust than, say, vector-space methods. When I started running LDA, I immediately discovered noise in my collection that had not previously been a problem. Running headers at the tops of pages, in particular, left traces: until I took out those headers, topics were suspiciously sensitive to the titles of volumes. But LDA is sensitive to noise, after all, because it is sensitive to everything else! On the whole, if you’re just fishing for interesting patterns in a large collection of documents, I think probabilistic techniques are the way to go.

Where to go next
The standard implementation of LDA is the one in MALLET. I haven’t used it yet, because I wanted to build my own version, to make sure I understood everything clearly. But MALLET is better. If you want a few examples of complete topic models on collections of 18/19c volumes, I’ve put some models, with R scripts to load them, in my github folder.

If you want to understand the technique more deeply, the first thing to do is to read up on Bayesian statistics. In this post, I gloss over the Bayesian underpinnings of LDA because I think the implementation (using a strategy called Gibbs sampling, which is actually what I described above!) is intuitive enough without them. And this might be all you need! I doubt most humanists will need to go further. But if you do want to tinker with the algorithm, you’ll need to understand Bayesian probability.

David Blei invented LDA, and writes well, so if you want to understand why this technique has “Dirichlet” in its name, his works are the next things to read. I recommend his Introduction to Probabilistic Topic Models. It recently came out in Communications of the ACM, but I think you get a more readable version by going to his publication page (link above) and clicking the pdf link at the top of the page.

Probably the next place to go is “Rethinking LDA: Why Priors Matter,” a really thoughtful article by Hanna Wallach, David Mimno, and Andrew McCallum that explains the “hyperparameters” I glossed over in a more principled way.

Then there are a whole family of techniques related to LDA — Topics Over Time, Dynamic Topic Modeling, Hierarchical LDA, Pachinko Allocation — that one can explore rapidly enough by searching the web. In general, it’s a good idea to approach these skeptically. They all promise to do more than LDA does, but they also add additional assumptions to the model, and humanists are going to need to reflect carefully about which assumptions we actually want to make. I do think humanists will want to modify the LDA algorithm, but it’s probably something we’re going to have to do for ourselves; I’m not convinced that computer scientists understand our problems well enough to do this kind of fine-tuning.

CVPR 2012 papers

丕子 — Thu, 05 Apr 2012 07:34:57 +0800

很久没有读paper了，唉。

http://cvpapers.com/cvpr2012.html

http://www.cvpr2012.org/program-details/papers

Orals

Title	Author(s)
Micro Phase Shifting	Mohit Gupta; Shree Nayar
On Multiple Foreground Cosegmentation	Gunhee Kim; Eric Xing
Face detection, pose estimation, and landmark localization in the wild	Xiangxin Zhu; Deva Ramanan
Supervised Hashing with Kernels	Wei Liu; Jun Wang; Rongrong Ji; Yu-Gang Jiang; Shih-Fu Chang
A Closed-Form Solution to Uncalibrated Photometric Stereo via Diffuse Maxima	Thoma Papadhimitri; Paolo Favaro
Camera Spectral Sensitivity Estimation from a Single Image under Unknown Illumination by using Fluorescence	Shuai Han; Yasuyuki Matsushita; Imari Sato; Takahiro Okabe; Yoichi Sato
Street-to-Shop: Cross-Scenario Clothing Retrieval via Parts Alignment and Auxiliary Set	Si Liu; Zheng Song; Guangcan LIU; Changsheng Xu; Hanqing Lu; Shuicheng YAN
Detecting activities of daily living in first-person camera views	Hamed Pirsiavash; Deva Ramanan
Weakly Supervised Structured Output Learning for Semantic Segmentation	Alexander Vezhnevets; Vittorio Ferrari; Joachim Buhmann
Progressive Graph Matching: Making a Move of Graphs via Probabilistic Voting	Minsu Cho; Kyoung Mu Lee
From Pixels to Physics: Probabilistic Color De-rendering	Ying Xiong; Kate Saenko; Trevor Darrell; Todd Zickler
Discriminative Illumination: Per-Pixel Classification of Raw Materials based on Optimal Projections of Spectral BRDF	Jinwei Gu; Chao Liu
Contextual Boost for Pedestrian Detection	Yuanyuan Ding; Jing Xiao
Face Alignment by Explicit Shape Regression	Xudong Cao; Yichen Wei; Fang Wen; Jian Sun
Accidental images within indoor scenes	Antonio Torralba; Bill Freeman
On Template-Based Reconstruction from a Single View: Analytical Solutions and Proofs of Well-Posedness for Developable, Isometric and Conformal Surfaces	Adrien Bartoli; Yan G¨¦rard; François Chadebecq
Jigsaw Puzzles with Pieces of Unknown Orientation	Andrew Gallagher
Understanding Collective Crowd Behaviors:Learning Mixture Model of Dynamic Pedestrian-Agents	Bolei Zhou; Xiaogang Wang
Learning Rotation-Aware Features: From Invariant Priors to Equivariant Descriptors	Uwe Schmidt; Stefan Roth
Are we ready for Autonomous Driving?	Andreas Geiger; Philip Lenz; Raquel Urtasun
General and Nested Wiberg Minimization	Dennis Strelow
Decomposing Global Light Transport using Time of Flight Imaging	Di Wu; Ramesh Raskar; Amit Agrawal; Andreas Velten; Matthew O'Toole
Curvature-Based Regularization for Surface Approximation	Carl Olsson; Yuri Boykov
Max-Margin Early Event Detectors	Minh Hoai; Fernando DelaTorre
Geodesic Flow Kernel for Unsupervised Domain Adaptation	Boqing Gong; Yuan Shi; Fei Sha; Kristen Grauman
Teaching 3D Geometry to Deformable Part Models	Bojan Pepik; Michael Stark; Peter Gehler; Bernt Schiele
Surface Regions of Interest for Viewpoint Selection	George Leifman; Elizabeth Shtrom; Ayellet Tal
The Shape Boltzmann Machine: a Strong Model of Object Shape	Seyed Mohammadali Eslami; Nicolas Heess; John Winn
Query Sensitive Binary Code Ranking	Xiao Zhang; Lei Zhang; Heung-Yeung Shum
A-Optimal Non-negative Projection for Image Representation	Zheng Yang; Haifeng Liu
Finding Animals: Semantic Segmentation using Regions and Parts	Pablo Arbelaez; Bharath Hariharan; Chunhui Gu; Saurabh Gupta; Lubomir Bourdev; Jitendra Malik
A Unified Approach to Salient Object Detection via Low Rank Matrix Recovery	Xiaohui Shen; Ying Wu
Pedestrian detection at 100 frames per second	Rodrigo Benenson; Markus Mathias; Radu Timofte; Luc Gool
Video from Nearly Still: an Application to Low Frame-rate Gait Recognition	Naoki Akae; Al Mansur; Yasushi Makihara; Yasushi Yagi
Stream-based Joint Exploration-Exploitation Active Learning	Chen Change Loy; Timothy Hospedales; Tao Xiang; Shaogang Gong
Affine-Invariant, Elastic Shape Analysis of Planar Contours	Darshan Bryner; Eric Klassen; Anuj Srivastava
Robust Visual Tracking via Multi-Task Sparse Learning	Tianzhu Zhang; Bernard Ghanem; Narendra Ahuja
Discriminative Virtual Views for Cross-View Action Recognition	Ruonan Li; Todd Zickler
Nonrigid Structure-from-Motion Factorization Made Easy	Yuchao Dai; Hongdong Li; Mingyi He
Autonomous Cleaning of Corrupted Scanned Documents - A Generative Modeling Approach	Zhenwen Dai; Jörg L¨¹cke
Stochastic Gradient on the Grassmannian for Online Foreground and Background Separation in Subsampled Video	Laura Balzano; Arthur Szlam; Jun He
Beyond Spatial Pyramids: Receptive Field Learning for Pooled Image Features	Yangqing Jia; Chang Huang; Trevor Darrell
A Theory of Flat Refractive Geometry	Amit Agrawal; Visesh Chari; Srikumar Ramalingam; Yuichi Taguchi
An Online Learned CRF Model for Multi-Target Tracking	Bo Yang; Ram Nevatia
What Are We Looking For: Towards Statistical Modeling of Saccadic Eye Movements and Visual Saliency	Xiaoshuai Sun; Hongxun Yao; Rongrong Ji; xianming Liu; Pengfei Xu
We are not Contortionist: Coupled Adaptive Learning for Head and Body Orientation Estimation in Surveillance Video	Cheng Chen; Jean-Marc Odobez
Connected Contours: a Contour Completion Model that respects Closure-Effect	Yansheng Ming; Hongdong Li; Xuming He
Large-scale image classification with trace-norm regularization	Zaid Harchaoui; Matthijs Douze; Mattis Paulin; Miko Dudik; Jerome Malick

3D交互界面

丕子 — Fri, 30 Mar 2012 07:58:01 +0800

技术飞速发展，说不定以后就可以用个眼镜就可以了。

本文转自：http://www.ifanr.com/80441

计算机从键鼠控制到触屏交互是个重大的进步，这使得人机交互更接近现实，降低了学习的门栏，使它进一步走向大众。那么，随着 3D 技术的不断进步，是否会有更好的方式取代触屏交互呢？毕竟现实中的交互都是三维方式的。

实际上，的确有一些研究者已经开始考虑这种可能性，近期曾有网站讲到了苹果的 3D 界面专利，证明苹果公司也在考虑 3D 化的交互界面。3D 界面的普及可能还有很长的路要走，但是目前已经有了一些实验性的成果。

BGR 报道了一种新的 3D 交互方式，由MIT 大学一位博士研究生（他同时也是 MIT 媒体实验室的研究助手） Jinha Lee 和微软的应用科学小组合作开发。值得关注的是，他不仅仅是展示了 3D 界面，同时也展示了一种全新的交互方式。

他开发的计算机有一个透明的 3D 屏幕，整个界面以3D 的方式展现，窗口和文件被放在屏幕之后的虚拟板之上。你可以如同现实中一样，直接用手势操作其中的物品。看起来，这比传统的触屏界面还要真实和直观。屏幕上有一个前置摄像头，可以跟踪用户的头部动作，对界面做出调整。用户也可以使用键盘和触控版来操作它。要了解它的神奇之处，可以看看下面这段视频：

在 Lee 的博客中写道：“计算机科学家 Ivan Sutherland 曾经将计算机屏幕称作‘通向数学乐园的镜子’，我经常期望着能够走进这个乐园，和其中的抽象物品交互。”他和微软的应用科学组合作创造的计算机，正好能够完成他的愿望。

Lee 并没有说他的研究变成实际产品的可能性。不过，随着 3D 技术的不断发展，类似的产品很可能在不久出现，这是否会成为人机交互的又一大进步呢？

您可能也喜欢：
最简单的上Twitter与Facebook之方法	Twitter预测股市行情？	[高清]Facebook, Twitter, YouTube, MySpace, Flikr 和 LinkedIn	推荐无处不在-看看Twitter	谷歌将收购Twitter？谷歌想收购Twitter？
无觅

您可能也喜欢：
十大最具智慧毕业典礼演讲	自己毕业设计用到的辅助工具	JOB:我们的征途是星辰大海—百度说吧招聘	百度竟然推出图像相似检索了：百度识图	百度也将模仿一下“哼唱搜索”？谷歌呢？
无觅

您可能也喜欢：
The Spider：一个“面向对象”的机器学习Matlab工具箱	又三个机器学习算法软件包：SVM、神经网络、流行学习等	[分享] 机器视觉工具箱-Machine Vision Toolbox for Matlab	记录基本统计学的书籍	Matlab统计工具箱：Statistics Toolbox
无觅

您可能也喜欢：
Histograms of Oriented Gradients (HOG)理解和源码	Mahout 0.3: 机器学习开源项目	UCI机器学习库和一些相关算法	PMTK3 :probabilistic modeling toolkit for Matlab/Octave	发现了几个机器学习的工具算法源码
无觅

您可能也喜欢：
[Job] 腾讯研究院实习职位	一年一年又一年	把女友升级为老婆时发生的BUG(z)	偶尔看见这么几句话，竟然为之所动	今明两天做苦力，阅卷工作
无觅

丕子

研究：Twitter情感趋势反映Facebook IPO走势

相关文章

毕业临近，旅旅游

相关文章

PMTK3 :probabilistic modeling toolkit for Matlab/Octave

Book Demos

Other Demos

相关文章

一个库，一个工具箱

from：http://people.csail.mit.edu/dhlin/softwares.html

Basic Computation Supporting Library (BCSLib)

Code website: http://code.google.com/p/bcslib/

Statistical Modeling and Inference Toolbox

Code website: http://code.google.com/p/smitoolbox/

相关文章

人生苍茫，最近状态

相关文章

Accepted Papers of ICMR 2012

Tentative schedule:

Accepted Papers:

Sponsors

相关文章

最近单曲循环--《再见青春》

最热文章

收藏：Topic modeling made just simple enough

Topic modeling made just simple enough.

相关文章

CVPR 2012 papers

Orals

相关文章

3D交互界面

相关文章

您可能也喜欢：
Google新图像检索发布视频	再次关注Google图像检索–嵌入广告	又一个图像检索引擎：PicsLikeThat	图像查询的特点(Characteristics of image queries)	再收集一下基于内容图像检索系统
无觅

您可能也喜欢：
爱就爱了-逝去的陈琳	80后最大“悬案”：扔纸飞机为啥先哈口气	金牌栏目危机四伏《同一首歌》唱九年终说再见	马克西姆-《野蜂飞舞》《出埃及记》《克罗地亚狂想曲》	看看这个机器人是怎么learn to create language
无觅

您可能也喜欢：
收集下2010之前的“基于LDA的Topic Model变形”的论文	LDA,咱们一起来学习	线性判别分析（LDA）, 主成分分析(PCA)	判别模型和生成模型(Discriminative Model& Generative Model)	Mallet：自然语言处理工具包
无觅

您可能也喜欢：
Tomasz：cvpr 2011: highlights from day 1	CVPR 2011 Best Paper-优秀论文	Conference Paper Awards—CVPR2010	CVPR2011录取结果	Jingen Liu一人独中两篇CVPR2011的Oral
无觅

您可能也喜欢：
[视频] 微软Surface新技术：移动3D界面	迪斯尼公开3D纹理映射源代码	[技术前瞻] 未来人机交互方式	[技术欣赏] 电子动态窗户和Google Earth 3D纽约	在自己的站点中嵌入3D立体相册
无觅