I wonder how do we evaluate feature detection/extraction methods (SIFT,SURF,MSER...) for object detection and tracking like pedestrians, lane vehicles etc.. Are there standard metrics for comparison? I have read blogs like http://computer-vision-talks.com/2011/07/comparison-of-the-opencvs-feature-detection-algorithms-ii/ some research papers like this. The problem is the more I learn the more I am confused.
It is very hard to estimate feature detectors per se, because features are only computation artifacts and not things that you are actually searching in images. Feature detectors do not make sense outside their intended context, which is affine-invariant image part matching for the descriptors that you have mentioned.
The very first usage of SIFT, SURF, MSER was multi-view reconstruction and automatic 3D reconstruction pipe-lines. Thus, these features are usually assessed from the quality of the 3D reconstrucution or image part matching that they provide. Roughly speaking, you have a pair of images that are related by a known transform (an affinity or an homography) and you measure the difference between the estimated homography (from the feature detector) and the real one.
This is also the method used in the blog post that you quote by the way.
In order to assess the practical interest of a detector (and not only its precision in an ideal multi-view pipe-line) some additional measurements of stability (under geometric and photometric changes) were added: does the number of detected features vary, does the quality of the estimated homography vary, etc.
Accidentally, it happens that these detectors may also work (also it was not their design purpose) for object detection and track (in tracking-by-detection cases). In this case, their performance is classically evaluated from more-or-less standardized image datasets, and typically expressed in terms of precision (probability of good answer, linked to the false alarm rate) and recall (probability of finding an object when it is present). You can read for example Wikipedia on this topic.
Addendum: What exactly do I mean by accidentally?
Well, as written above, SIFT and the like were designed to match planar and textured image parts. This is why you always see example with similar images from a dataset of graffiti.
Their extension to detection and tracking was then developed in two different ways:
While doing multiview matching (with a spherical rig), Furukawa and Ponce built some kind of 3D locally-planar object model, that they applied then to object detection in presence of severe occlusions. This worlk exploits the fact that an interesting object is often locally planar and textured;
Other people developed a less original (but still efficient in good conditions) approach by considering that they had a query image of the object to track. Individual frame detections are then performed by matching (using SIFT, etc.) the template image with the current frame. This exploits the fact that there are few false matchings with SIFT, that objects are usually observed in a distance (hence are usually almost planar in images) and that they are textured. See for example this paper.
Related
PREMISE:
I'm really new to Computer Vision/Image Processing and Machine Learning (luckily, I'm more expert on Information retrieval), so please be kind with this filthy peasant! :D
MY APPLICATION:
We have a mobile application where the user takes a photo (the query) and the system returns the most similar picture thas was previously taken by some other user (the dataset element). Time performances are crucial, followed by precision and finally by memory usage.
MY APPROACH:
First of all, it's quite obvious that this is a 1-Nearest Neighbor problem (1-NN). LSH is a popular, fast and relatively precise solution for this problem. In particular, my LSH impelementation is about using Kernalized Locality Sensitive Hashing to achieve a good precision to translate a d-dimension vector to a s-dimension binary vector (where s<<d) and then use Fast Exact Search in Hamming Space
with Multi-Index Hashing to quickly find the exact nearest neighbor between all the vectors in the dataset (transposed to hamming space).
In addition, I'm going to use SIFT since I want to use a robust keypoint detector&descriptor for my application.
WHAT DOES IT MISS IN THIS PROCESS?
Well, it seems that I already decided everything, right? Actually NO: in my linked question I face the problem about how to represent the set descriptor vectors of a single image into a vector. Why do I need it? Because a query/dataset element in LSH is vector, not a matrix (while SIFT keypoint descriptor set is a matrix). As someone suggested in the comments, the commonest (and most efficient) solution is using the Bag of Features (BoF) model, which I'm still not confident with yet.
So, I read this article, but I have still some questions (see QUESTIONS below)!
QUESTIONS:
First and most important question: do you think that this is a reasonable approach?
Is k-means used in the BoF algorithm the best choice for such an application? What are alternative clustering algorithms?
The dimension of the codeword vector obtained by the BoF is equal to the number of clusters (so k parameter in the k-means approach)?
If 2. is correct, bigger is k then more precise is the BoF vector obtained?
There is any "dynamic" k-means? Since the query image must added to the dataset after the computation is done (remember: the dataset is formed by the images of all submitted queries) the cluster can change in time.
Given a query image, is the process to obtain the codebook vector the same as the one for a dataset image, e.g. we assign each descriptor to a cluster and the i-th dimension of the resulting vector is equal to the number of descriptors assigned to the i-th cluster?
It looks like you are building codebook from a set of keypoint features generated by SIFT.
You can try "mixture of gaussians" model. K-means assumes that each dimension of a keypoint is independent while "mixture of gaussians" can model the correlation between each dimension of the keypoint feature.
I can't answer this question. But I remember that the SIFT keypoint, by default, has 128 dimensions. You probably want a smaller number of clusters like 50 clusters.
N/A
You can try Infinite Gaussian Mixture Model or look at this paper: "Revisiting k-means: New Algorithms via Bayesian Nonparametrics" by Brian Kulis and Michael Jordan!
Not sure if I understand this question.
Hope this help!
I've been playing with the excellent GPUImage library, which implements several feature detectors: Harris, FAST, ShiTomas, Noble. However none of those implementations help with the feature extraction and matching part. They simply output a set of detected corner points.
My understanding (which is shakey) is that the next step would be to examine each of those detected corner points and extract the feature from then, which would result in descriptor - ie, a 32 or 64 bit number that could be used to index the point near to other, similar points.
From reading Chapter 4.1 of [Computer Vision Algorithms and Applications, Szeliski], I understand that using a BestBin approach would help to efficient find neighbouring feautures to match, etc. However, I don't actually know how to do this and I'm looking for some example code that does this.
I've found this project [https://github.com/Moodstocks/sift-gpu-iphone] which claims to implement as much as possible of the feature extraction in the GPU. I've also seen some discussion that indicates it might generate buggy descriptors.
And in any case, that code doesn't go on to show how the extracted features would be best matched against another image.
My use case if trying to find objects in an image.
Does anyone have any code that does this, or at least a good implementation that shows how the extracted features are matched? I'm hoping not to have to rewrite the whole set of algorithms.
thanks,
Rob.
First, you need to be careful with SIFT implementations, because the SIFT algorithm is patented and the owners of those patents require license fees for its use. I've intentionally avoided using that algorithm for anything as a result.
Finding good feature detection and extraction methods that also work well on a GPU is a little tricky. The Harris, Shi-Tomasi, and Noble corner detectors in GPUImage are all derivatives of the same base operation, and probably aren't the fastest way to identify features.
As you can tell, my FAST corner detector isn't operational yet. The idea there is to use a lookup texture based on a local binary pattern (why I built that filter first to test the concept), and to have that return whether it's a corner point or not. That should be much faster than the Harris, etc. corner detectors. I also need to finish my histogram pyramid point extractor so that feature extraction isn't done in an extremely slow loop on the GPU.
The use of a lookup texture for a FAST corner detector is inspired by this paper by Jaco Cronje on a technique they refer to as BFROST. In addition to using the quick, texture-based lookup for feature detection, the paper proposes using the binary pattern as a quick descriptor for the feature. There's a little more to it than that, but in general that's what they propose.
Feature matching is done by Hamming distance, but while there are quick CPU-side and CUDA instructions for calculating that, OpenGL ES doesn't have one. A different approach might be required there. Similarly, I don't have a good solution for finding a best match between groups of features beyond something CPU-side, but I haven't thought that far yet.
It is a primary goal of mine to have this in the framework (it's one of the reasons I built it), but I haven't had the time to work on this lately. The above are at least my thoughts on how I would approach this, but I warn you that this will not be easy to implement.
For object recognition / these days (as of a couple weeks ago) best to use tensorflow /Convolutional Neural Networks for this.
Apple has some metal sample code recently added. https://developer.apple.com/library/content/samplecode/MetalImageRecognition/Introduction/Intro.html#//apple_ref/doc/uid/TP40017385
To do feature detection within an image - I draw your attention to an out of the box - KAZE/AKAZE algorithm with opencv.
http://www.robesafe.com/personal/pablo.alcantarilla/kaze.html
For ios, I glued the Akaze class together with another stitching sample to illustrate.
detector = cv::AKAZE::create();
detector->detect(mat, keypoints); // this will find the keypoints
cv::drawKeypoints(mat, keypoints, mat);
// this is the pseudo SIFT descriptor
.. [255] = {
pt = (x = 645.707153, y = 56.4605064)
size = 4.80000019
angle = 0
response = 0.00223364262
octave = 0
class_id = 0 }
https://github.com/johndpope/OpenCVSwiftStitch
Here is a GPU accelerated SIFT feature extractor:
https://github.com/lukevanin/SIFTMetal
The code is written in Swift 5 and uses Metal compute shaders for most operations (scaling, gaussian blur, key point detection and interpolation, feature extraction). The implementation is largely based on the paper and code from the "Anatomy of the SIFT Method Article" published in the Image Processing Online Journal (IPOL) in 2014 (http://www.ipol.im/pub/art/2014/82/). Some parts are based on code by Rob Whess (https://github.com/robwhess/opensift), which I believe is now used in OpenCV.
For feature matching I tried using a kd-tree with the best-bin first (BBF) method proposed by David Lowe. While BBF does provide some benefit up to about 10 dimensions, with a higher number of dimensions such as used by SIFT, it is no better than quadratic search due to the "curse of dimensionality". That is to say, if you compare 1000 descriptors against 1000 other descriptors, it stills end up making 1,000 x 1,000 = 1,000,000 comparisons - the same as doing brute-force pairwise.
In the linked code I use a different approach optimised for performance over accuracy. I use a trie to locate the general vicinity for potential neighbours, then search a fixed number of sibling leaf nodes for the nearest neighbours. In practice this matches about 50% of the descriptors, but only makes 1000 * 20 = 20,000 comparisons - about 50x faster and scales linearly instead of quadratically.
I am still testing and refining the code. Hopefully it helps someone.
If the data to cluster are literally points (either 2D (x, y) or 3D (x, y,z)), it would be quite intuitive to choose a clustering method. Because we can draw them and visualize them, we somewhat know better which clustering method is more suitable.
e.g.1 If my 2D data set is of the formation shown in the right top corner, I would know that K-means may not be a wise choice here, whereas DBSCAN seems like a better idea.
However, just as the scikit-learn website states:
While these examples give some intuition about the algorithms, this
intuition might not apply to very high dimensional data.
AFAIK, in most of the piratical problems we don't have such simple data. Most probably, we have high-dimensional tuples, which cannot be visualized like such, as data.
e.g.2 I wish to cluster a data set where each data is represented as a 4-D tuple <characteristic1, characteristic2, characteristic3, characteristic4>. I CANNOT visualize it in a coordinate system and observes its distribution like before. So I will NOT be able to say DBSCAN is superior to K-means in this case.
So my question:
How does one choose the suitable clustering method for such an "invisualizable" high-dimensional case?
"High-dimensional" in clustering probably starts at some 10-20 dimensions in dense data, and 1000+ dimensions in sparse data (e.g. text).
4 dimensions are not much of a problem, and can still be visualized; for example by using multiple 2d projections (or even 3d, using rotation); or using parallel coordinates. Here's a visualization of the 4-dimensional "iris" data set using a scatter plot matrix.
However, the first thing you still should do is spend a lot of time on preprocessing, and finding an appropriate distance function.
If you really need methods for high-dimensional data, have a look at subspace clustering and correlation clustering, e.g.
Kriegel, Hans-Peter, Peer Kröger, and Arthur Zimek. Clustering high-dimensional data: A survey on subspace clustering, pattern-based clustering, and correlation clustering. ACM Transactions on Knowledge Discovery from Data (TKDD) 3.1 (2009): 1.
The authors of that survey also publish a software framework which has a lot of these advanced clustering methods (not just k-means, but e.h. CASH, FourC, ERiC): ELKI
There are at least two common, generic approaches:
One can use some dimensionality reduction technique in order to actually visualize the high dimensional data, there are dozens of popular solutions including (but not limited to):
PCA - principal component analysis
SOM - self-organizing maps
Sammon's mapping
Autoencoder Neural Networks
KPCA - kernel principal component analysis
Isomap
After this one goes back to the original space and use some techniques that seems resonable based on observations in the reduced space, or performs clustering in the reduced space itself.First approach uses all avaliable information, but can be invalid due to differences induced by the reduction process. While the second one ensures that your observations and choice is valid (as you reduce your problem to the nice, 2d/3d one) but it loses lots of information due to transformation used.
One tries many different algorithms and choose the one with the best metrics (there have been many clustering evaluation metrics proposed). This is computationally expensive approach, but has a lower bias (as reducting the dimensionality introduces the information change following from the used transformation)
It is true that high dimensional data cannot be easily visualized in an euclidean high dimensional data but it is not true that there are no visualization techniques for them.
In addition to this claim I will add that with just 4 features (your dimensions) you can easily try the parallel coordinates visualization method. Or simply try a multivariate data analysis taking two features at a time (so 6 times in total) to try to figure out which relations intercour between the two (correlation and dependency generally). Or you can even use a 3d space for three at a time.
Then, how to get some info from these visualizations? Well, it is not as easy as in an euclidean space but the point is to spot visually if the data clusters in some groups (eg near some values on an axis for a parallel coordinate diagram) and think if the data is somehow separable (eg if it forms regions like circles or line separable in the scatter plots).
A little digression: the diagram you posted is not indicative of the power or capabilities of each algorithm given some particular data distributions, it simply highlights the nature of some algorithms: for instance k-means is able to separate only convex and ellipsoidail areas (and keep in mind that convexity and ellipsoids exist even in N-th dimensions). What I mean is that there is not a rule that says: given the distributiuons depicted in this diagram, you have to choose the correct clustering algorithm consequently.
I suggest to use a data mining toolbox that lets you explore and visualize the data (and easily transform them since you can change their topology with transformations, projections and reductions, check the other answer by lejlot for that) like Weka (plus you do not have to implement all the algorithms by yourself.
In the end I will point you to this resource for different cluster goodness and fitness measures so you can compare the results rfom different algorithms.
I would also suggest soft subspace clustering, a pretty common approach nowadays, where feature weights are added to find the most relevant features. You can use these weights to increase performance and improve the BMU calculation with euclidean distance, for example.
When detecting objects using SURF, how can a plot a graph for false positives and hits using the Good matches and several keypoints?
(A) How do I get the statistics of good matches i.e an ROC plot or the true positives vs false positives of detection from so many of the line descriptors?Can somebody put a code for plotting true positves vs false positive statistics.
(B)**Secondly,there are many resources vdo1 , vdo2and implemetations, papers ( Object tracking using improved Camshift with SURF method ;
A Study on Moving Object Tracking Algorithm Using SURF Algorithm
and Depth Information
) which say that SURF and SIFT can be used for tracking in combination with camshift or meanshift.
But, what I fail to understand is that we need prediction algorithm like Kalman filters or tracking algorithm like Camshift,mean shift or template differencing(not sure) for tracking.So,how come some video implementations and tutorial say that Lukas Kanade Optical flow,SIFT,SURF is tracking objects whereas the papers mention to club either camshift or meanshift.Am I missing out on some conceptual matter?
Shall be obliged for pointers and a detailed explanation on how SURF or SIFT or feature based methods can be used for tracking alone?
Lucas-Kandae with pyramid (pyrLK) is a method that looks for a small shift in a single feature location. It can do this to many features at once. Camshift and meanshift track a statistic for a group of features. You can also just try to use a matcher, to find where the features went on the next frame. GoodFeturesToTrack, SIFT and SURF are algorithms that find points that should be easy to find and tell apart one from another. SURF and SIFT include also descriptors, that characterise those features in a way which can ignore size change, orientation change or both.
Kalman filter is used to refine Your results. It is able to shrink the area where the answer should lay, because algorithms above are not perfect.
As for the code, I haven't done too much tracking except Shi-Thomasi + pyrLK, so I dont't think I can help.
I am wondering how can I claim that I correctly catch the "noise" in my data ?
To be more specific, take Principle Component Analysis as example, we know that in PCA, after doing SVD, we can zeros out the small singular values and reconstruct the original matrix using low-rank approximation.
Then can I claim what's been ignored is indeed noise in the data ?
Is there any evaluation metric for this ?
The only method I can come up with is simply subtract the original data from the reconstructed data.
Then, try to fit a Gaussian over it, seeing if the fitness is good.
Is that conventional method in field like DSP ??
BTW, I think in typical machine learning tasks, the measurement would be the follow up classification performance, but since I am doing purely generative model, there are no labels attached.
The way I see it, the definition of noise would depend on the domain of the problem. Therefore the strategy for reducing it would be different on each domain.
For instance, having a noisy signal in problems like seismic formation classification or a noisy image on a face classification problem would be drastically different to the noise produced by improperly tagged data in a medical diagnostic problem or the noise because similar words with different meaning in a language classification problem for documents.
When the noise is because of a given (or a set of) data point, then the solution is as simple as ignore those data points (although identify those data points most of the time is the challenging part)
From your example I guess you are more concerning about the case when the noise is embedded into the features (like in the seismic example). Sometimes people tend to pre-process the data with a noise reduction filter like the median filter (http://en.wikipedia.org/wiki/Median_filter). In contrast, some other people tend to reduce the dimension of the data to reduce noise, and PCA is used in this scenario.
Both strategies are valid, and normally people try both and cross-validate them to see which one gave better results.
What you did is a good metric to check gaussian noise. However, for non-gaussian noise your metric can give you false negatives (bad fitness but still good noise reduction)
Personally, if you want to prove the efficacy of the noise reduction, I'd use a task-based evaluation. I assume you're doing this for some purpose, to solve some problem? If so, solve the task with the original noisy matrix and the new clean one. If the latter works better, what was discarded was noise, for the purposes of the task you're interested in. I think some objective measure of noise is pretty hard to define.
I have found this. it is very resoureful, needs good time to understand.
https://sci2s.ugr.es/noisydata#Introduction%20to%20Noise%20in%20Data%20Mining