Hi Engineers & Developers!
Hope you are doing well. I am working with a "Forensic Watermarking" Solution and there is a process running which is called "Embedding 0/1". It embeds 0 or 1 into each video frames then encoded each frame. I am not clear actually what is happening inside the Embedder. Can you guys any idea ?
Regards
Tanmaya
The embedder is 'hiding' information in the image for a particular frame.
The actual technique to hide that data may be different from implementation to implementation - there is quite a bit of literature and research in this domain, Steganography.
There is a very good summary in a recent IEEE paper:
https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9187785
(In case link breaks in future, paper is called 'Digital Steganography and Watermarking for Digital Images: A Review of Current Research Directions
OLEG EVSUTIN 1,3, ANNA MELMAN2,
AND ROMAN MESHCHERYAKOV3, (Senior Member, IEEE)')
Related
I'm curious about how I may be able to more reliably recognise the value and the suit of playing card images. Here are two examples:
There may be some noise in the images, but I have a large dataset of images that I could use for training (roughly 10k pngs, including all values & suits).
I can reliably recognise images that I've manually classified, if I have a known exact-match using a hashing method. But since I'm hashing images based on their content, then the slightest noise changes the hash and results in an image being treated as unknown. This is what I'm looking to reliably address with further automation.
I've been reviewing the 3.05 documentation on training tesseract:
https://github.com/tesseract-ocr/tesseract/wiki/Training-Tesseract#automated-method
Can tesseract only be trained with images found in fonts? Or could I use it to recognise the suits for these cards?
I was hoping that I could say that all images in this folder correspond to 4c (e.g. the example images above), and that tesseract would see the similarity in any future instances of that image (regardless of noise) and also read that as 4c. Is this possible? Does anyone here have experience with this?
This has been my non-tesseract solution to this, until someone proves there's a better way. I've setup:
Caffe: http://caffe.berkeleyvision.org/install_osx.html
Digits: https://github.com/NVIDIA/DIGITS/blob/master/docs/BuildDigits.md
Getting these to running was the hardest part. Next, I used my dataset to train a new caffe network. I prepared my dataset into a single depth folder structure:
./card
./card/2c
./card/2d
./card/2h
./card/2s
./card/3c
./card/3d
./card/3h
./card/3s
./card/4c
./card/4d
./card/4h
./card/4s
./card/5c
./card/5d
./card/5h
./card/5s
./card/6c
./card/6d
./card/6h
./card/6s
./card/7c
./card/7d
./card/7h
./card/7s
./card/8c
./card/8d
./card/8h
./card/8s
./card/9c
./card/9d
./card/9h
./card/9s
./card/_noise
./card/_table
./card/Ac
./card/Ad
./card/Ah
./card/As
./card/Jc
./card/Jd
./card/Jh
./card/Js
./card/Kc
./card/Kd
./card/Kh
./card/Ks
./card/Qc
./card/Qd
./card/Qh
./card/Qs
./card/Tc
./card/Td
./card/Th
./card/Ts
Within Digits, I chose:
Datasets tab
New Dataset Images
Classification
I pointed it to my card folder, e.g: /path/to/card
I set the validation % to 13.0%, based on the discussion here: https://stackoverflow.com/a/13612921/880837
After creating the dataset, I opened the models tab
Chose my new dataset.
Chose the GoogLeNet under Standard Networks, and left it to train.
I did this several times, each time I had new images in the dataset. Each learning session took 6-10 hours, but at this stage I can use my caffemodel to programmatically estimate what each image is expected to be, using this logic:
https://github.com/BVLC/caffe/blob/master/examples/cpp_classification/classification.cpp
The results are either a card (2c, 7h, etc), noise, or table. Any estimates with an accuracy bigger than 90% are most likely correct. The latest run correctly recognised 300 out of 400 images, with only 3 mistakes. I'm adding new images to the dataset and retraining the existing model, further tuning the result accuracy. Hope this is valuable to others!
While I wanted the high level steps here, this was all done with large thanks to David Humphrey and his github post, I really recommend reading it and trying it out if you're interested in learning more: https://github.com/humphd/have-fun-with-machine-learning
I am new to iSCSI programming. I have security camera videos written in iSCSI blocks in a SAN. I need to construct playback of the some video from those blocks. I have hard time finding any tutorial or lead to solve this issue.
Thanks in advance.
I'd start with the libiscsi tools. That should let you read the contents of the LUN from anywhere. Beyond that, there are no tutorials. Unless you've omitted details, you've basically lot a large block of data with video hidden in it somewhere.
Start by trying to figure out the structure of the contents of the LUN. If you're really lucky, it's just constantly rewriting a raw MPEG stream over the whole LUN, so you might copy some of the contents to a file and see what ffmpeg says about it. If you're unlikely, you've just got a bit file format reverse engineering problem on your hands, and that's far outside the scope of stack overflow, even if you gave us details. Some places to start looking:
http://en.wikibooks.org/wiki/Reverse_Engineering/File_Formats
http://www.sans.org/reading-room/whitepapers/forensics/reverse-engineering-microsoft-exfat-file-system-33274
https://hackaday.io/project/3149-reverse-engineering-design-file-formats
I'm having an idea to create a Censor Plugin/Extension for VLC Player..
Problem Scenario :
An Adult-Scene for 1 minute in a nice movie makes it not watchable with Family.
My Solution :
Create a Plugin/Extension which does the following
Reads time positions from a file similar to subtitle files
Skip these time positions (which are adult or inappropriate) when playing
Help i needed :
I searched in Google and in videolan website, But can't find an exact solution
Are there already similar Plugins available?
Where should i start?
Please help me if you could guys.. thanks..
Same looking for having/developing Exact same solution. This might be helpful to you.
http://code.google.com/p/movie-content-editor/
A similar thing is also available on github:
https://github.com/rdp/sensible-cinema
You may also want to read this discussion thread:
https://forum.videolan.org/viewtopic.php?t=89466
finding great similar answer here
If you chop random bytes out the movie is likely not playable. The player might crash or fail to resynchronize the stream – the video might just stop. Plus, you're gonna have a hard time figuring out where the "adult" bytes are, so to speak.
If you already know where the parts are that you want to cut out, I would edit the file in any of the numerous video editors. Even Windows Movie Maker or iMovie would do the job, and those are easily available on both major OSes.
This is a requested feature for VLC. Not really anything user-friendly out there. Still, VLC offers the possibility to create playlists in a certain format that would mute or skip parts of a file. This is called XSPF. You might be able to figure out the proper format for this.
Also, there's movie-content-editor:
A VLC based editor built in python that allows users to create and use custom filter files to make movies more family friendly. Allows users to have the player automatically mute specific words or skip certain scenes based on the content of those scenes.
And sensible-cinema:
Clean Editing Movie Player allows you watch edited movies by applying delete lists (EDL's) (i.e. "mute out" or "cut out" scenes) to DVD's/files, with preliminary support for also applying them to arbitrary web/internet based players like netflix instant, hulu/hulu plus etc
See also these threads on The VideoLAN Forums:
auto skip unwanted parts of a video
Clearplay-like (content filter) module exists?
I am looking for algorithms that allow text extraction from websites. I do not mean "strip html", or any of the hundreds of libraries that allow this.
So for example for a news article I would like to identify the heading and all the text, but not the comments section and so on.
Are there any algorithms for that out there? Thank you!
In computer science literature this problem is usually referred to as the page segmentation or boiler plate detection problem. See the report Boilerplate Detection using Shallow Text Features and its related blog post. Also, I have a few reports and software sites bookmarked that address the problem. Also, see this stackoverflow question.
there are a few open source tools available that do similar article extraction tasks.
https://github.com/jiminoc/goose which was open source by Gravity.com
It has info on the wiki as well as the source you can view. There are dozens of unit tests that show the text extracted from various articles.
"Content extraction" is a very difficult topic. There are no common standards to identify the "main-article" content (there are several approaches to make HTML easier readably for crawlers, e.g. schema.org, but none of these is very popularly used).
So it turns out, that if you want good results, its probably best to define your own XPath selectors for each (news) website you want to scrape. Although there are some APIs for HTML content extraction, but as I said its very hard to develop an algorithm which works for every site.
Some APIs you could use:
alchemyapi.com
diffbot.com
boilerpipe-web.appspot.com
aylien.com
textracto.com
What you're trying to do is called "content extraction". It turns out to be a surprisingly hard problem to solve well, and many naive solutions do quite badly.
Instapaper and Readability both have to solve this, and you may learn something from looking at their solutions. They also both provide services that you may be able to take advantage of - perhaps you can outsource your problem to them and let their API take care of it. :)
Failing that, a search for "html content extraction" returns a great deal of useful results, including a number of papers on the subject.
I compared a few different libraries, and had really great luck with Mozilla's Readability library (Node), or its Python wrapper.
For example, take this CNN article: https://edition.cnn.com/2022/06/01/tech/elon-musk-tesla-ends-work-from-home/index.html
Readability successfully returns only the relevant data:
New York (CNN Business) Elon Musk is demanding that Tesla office workers return to in-person work or leave the company. The policy, disclosed in leaked emails Musk sent to Tesla's executive staff Tuesday, was first reported by electric vehicle news site Electrek. "Anyone who wishes to do remote work must be in the office for a minimum (and I mean *minimum*) of 40 hours per week or depart Tesla. This is less than we ask of factory workers," Musk wrote, adding that the office must be the employee's primary workplace where the other workers they regularly interact with are based — "not a remote branch office unrelated to the job duties." Musk said he would personally review any request for exemption from the policy, but that for the most part, "If you don't show up, we will assume you have resigned."
etc.
I think your best shoot is study what information can you get from the metadata and write a good html parser, oEmbed could be a good standard =)
https://oembed.com/#section7
I've read about systems which use the Flickr database of photos to fill in gaps in photos (http://blogs.zdnet.com/emergingtech/?p=629).
How feasible is a system like this? I was toying with the idea (not just a way of killing time but as a good addition to something I am coding) of using Flickr to get photos of a certain entity (in this case, race tracks) and reconstruct a model. My biggest concern is that there aren't enough photos of a particular track and even then, it would be difficult to tell if two photos are of the same part of the racetrack, in which case one of them may be irrelevant.
How feasible is something like this? Is it worth attempting by a sole developer?
Sounds like you're wanting to build a Photosynth style system - check out Blaise Aguera y Arcas' demo at TED back in 2007. There's a section about 4 minutes in where he builds a model of the Sagrada Família from photographs.
I say +1 for photosynth answer, its a great tool. Not sure how well you could incorporate it into your own app though.
Its definately feasable. Anything is possible. And yes, doable for a single developer, just depends how much free time you have. It would be great to see something like this integrated into Virtual Earth or Google Maps Street View. Someone who could nail some software like this could help 3D model the entire world based purely on photographs. That would be a great product and make any single developer rich and famous.
So get coding. :)
I have plenty of free time, as I am in between jobs.
One way to do it is to get an overhead view of the track layout, make a blueprint based on this model, and then get one photo of the track and mimic the track's road colour. That would be a start.
LINQ to Flickr on codeplex has a great API and would be helpful for your task.