| Who | When |
Messages | |
|
|
|
travestia
|
22
|
 |
|
07-24-2008 03:42 AM ET (US)
|
|
|
| travestia
|
21
|
|
|
07-21-2008 01:30 AM ET (US)
|
|
|
| |
Messages 20-19 deleted by topic administrator between 07-21-2008 02:09 AM and 06-25-2008 02:26 AM |
| sunglow
|
18
|
|
|
06-13-2008 05:15 AM ET (US)
|
|
|
| cila
|
17
|
|
|
06-12-2008 04:16 AM ET (US)
|
|
|
| |
Messages 16-13 deleted by topic administrator between 05-16-2008 08:08 AM and 07-23-2006 02:04 AM |
| mcity4
|
12
|
|
|
11-29-2004 11:25 PM ET (US)
|
|
Marlboro $15.95with delivery at www.mcity4.com!
Famous brand cigarettes with delivery around the world !
Marlboro $15.95 / Kent $18.95
Camel $15.95 / B&H $14.95
WEST $14.95 / Captain Black $26.95
BOND $12.95 / M.Carlo $13.95
Davidoff $20.95 / Vogue $20.95
Magna $14.95 / More $14.95
Dunhill $21.95 / Rothmans $17.95
Cigaronne $29.95 / L&M $13.95
Winston $15.95 / Parliament $19.95
Dunhill International $22.95 /
Lucky Strike $15.95
www.mcity4.com www.mcity4.com www.mcity4.com
On-line cigarette shop !
|
| mcity4
|
11
|
|
|
09-08-2004 12:44 AM ET (US)
|
|
Famous brand cigarettes with delivery around the world!
Marlboro $13.95 / Kent $16.95
Camel $13.95 / 555 $17.95
WEST $12.95 / Pall Mall $12.95
BOND $12.95 / Prince $14.95
Davidoff $19.95 / Vogue $17.95
Magna $12.95 / More $12.95
Dunhill $21.95 / Rothmans $14.95
Cigaronne $31.95 / L&M $12.95
Winston $13.95 / Salem $14.95
Dunhill International $20.95 /
Gauloises Blondes $13.95 /
Gitanes Blondes $13.95 /
Parliament $17.95 / M.Carlo $12.95
www.mcity4.com www.mcity4.com www.mcity4.com
On-line cigarette shop !
|
| Omer Rashid
|
10
|
|
|
05-20-2004 07:18 PM ET (US)
|
|
hello All, hope that u all are fine ...Recognizing at a distance . I am doing my final project in the same descipline . Its recognition of human through gaits. Can any one suggest me a good techique for that . I am implementing . I am currently working on the backgroud removal and stuff like that and then I have planned to work on the object by taking the stick image of the object after removing the background. I want u all to suggest me some thing good for that to recognize or the distinguishing features of human in gaits which distinguishes them from other . I shall be very thankful to u . Bye . Omer
|
| Shinko Cheng
|
9
|
|
|
10-30-2003 10:16 PM ET (US)
|
|
Edited by author 10-30-2003 10:39 PM
Optical flow is relatively appearance invariant. The implications of using it to recognize certain motions is very interesting. I hope there can be some extensive flow models of various human motions, creating a library with which to perform recognition tasks.
The work reminds me of the randomized looping of video sequences, e.g. taking a sequence of a candle flame and looping it over and over again at random places. In it, they also used the similarity matrix but on full image frames.
This paper claims that the novel contribution is the motion descriptor. There appears to be however so much other work on the side for it. It appears to me that this person spent a lot of time on presenting its promise: skelton transfer, action synthesis, figure correction, "action database", if these things were indeed not research contributions themselves as the author seems to imply. The action synthesis one certainly doesn't sound too new; candle flame looping uses a similar concept confusion matrix concept. I guess what i'm saying is a follow up paper is called for on all the "promise" made in the paper.
|
| Sunny Chow
|
8
|
|
|
10-23-2003 03:13 AM ET (US)
|
|
Mike,
You're right, I misspoke a bit when I said it was a weakness..
One of the algorithm's applications was to remove occlusion (which suggests it will handle limited occlusion okay), but I doubt it will be able to handle much more than a simple head bobbing into view.
As for backgrounds, as long as the background is not moving relative to the subject, I think all will be good. However, if the subject was running around with a busy background, the motion induced by the background may throw off the classification.
|
| Mike McCracken
|
7
|
|
|
10-23-2003 02:33 AM ET (US)
|
|
The fact that you need to have previous data to match with doesn't seem like such a weakness to me. I expect that in most applications, previous data wouldn't be too hard to come by, and something as simple as flagging "unclassifiable action" for human review would be good enough to cover cases where the subjects being monitored began to levitate, for instance...
The discussion about how the examples all have static backgrounds reminded me of the Three Brown Mice project (from Kristin and Serge) - I was thinking that if a football player were running next to another player, the scheme from the paper would break down. However, in the mice project, they were able to classify actions despite occlusions (if I recall correctly) It seems like you'd have to do more for this to work with different backgrounds, and some backgrounds might just be too hard..
|
| Sunny Chow
|
6
|
|
|
10-23-2003 01:16 AM ET (US)
|
|
That's one of the weaknesses of this algorithm; that you have to experience it to be able to classify it. This means that unless the jumping, or some other random action was included in some sequence stored in a database already, this algorithm won't be able to classify it.
Stabilization is just one of the many assumptions the paper makes, I don't think it tries at all to do it...
|
| Diem Vu
|
5
|
|
|
10-23-2003 12:14 AM ET (US)
|
|
Very interesting paper, but it seems obviously failed to detect jump or similar actions unless they have some tricks in the stabilized proccess.
|
| Meifang Huang
|
4
|
|
|
10-22-2003 11:37 PM ET (US)
|
|
This paper uses optical flows to extract the motion features from sequential frames, and compares different sequences using the correlation of these features. It is a simple method but works quite well. It compares the motion descriptors in both the spatial and temporal dimension and also solves the problem when the motion rates are different in two sequences. The synthesis part of "Do as I Do" and "Do as I Say" sounds fancy, I am looking forward to seeing the demonstration.
|
| Neil Alldrin
|
3
|
|
|
10-22-2003 10:22 PM ET (US)
|
|
I also like this paper. Comparing optical flows as a similarity measure is a pretty cool idea. One thing to be wary of (as Serge pointed out to me today) is that the background is relatively constant in most of their videos. If this were not so, I imagine there would be a lot more noise in their motion descriptors, which could lead to a breakdown of their classification mechanism.
|
| Jing Shiau
|
2
|
|
|
10-22-2003 09:28 PM ET (US)
|
|
I like this paper. =) Using optical flow to get a motion descriptor isn't exactly intuitive, but once the idea is introduced, it is fairly straight forward.
Even though the authors say the main contribution of the paper is in the motion descriptor, I like their idea of "Do as I do" and "Do as I say". Maybe this can be used to creat personalized avatars in virtual reality applications?
At first I thought this paper was trying to recognize action just like the "Recognition of human gaits" paper, but seem that they are trying to do different things. The resolution of the figures used in this paper are probably too low to apply the technique of representing human skeleton as a kinematic chain.
|
| Matt Clothier
|
1
|
|
|
10-19-2003 07:08 PM ET (US)
|
|
I remember reading this paper. ;) It is cool to think that a person only 30 pixels tall can provide sufficient information to determine the person's action (and then be able to mimic it with "do as I do"). This paper also reminds me a lot of the "Recognition of human gaits" paper that we looked at recently.
I do wonder about the skeletal transfer system. By using either the hand-marked joint locations or the 3D motion capture data, these are databases that are used outside the context of the scene. It seems to me that it would be better to build a training set using the actual video sequence itself. It may be that they decided not to do this because the resolution of a person is not high enough to determine the exact action. Does anyone think that using the "human skeleton represented as a kinematic chain" (from the "Recognition of human gaits" paper) would work here?
|
