Scraawl PixL Help Desk

Frequently Asked QuestionsFrequently Asked Questions

Don’t see your question below? Contact us!
- - -
1. I created a new account but I still can’t log in, why?

After registering, you should have received an e-mail confirming your registration. Open that e-mail and click on the provided link to complete the registration process. If you have not received the e-mail, please check your spam folder or contact us to help resolve the issue.

2. Scraawl does not accept my password. What should I do?

Make sure your password is at least 8 characters long and contains at least one letter, one number, and one of the following special characters: ! @ * # $ _% ^

3. I forgot my password. What should I do?

Use the Forgot password? link on the Scraawl home page, and follow the instructions to reset your password.

4. How do I close or delete my Scraawl account?

Go to your My Profile page and click on the Delete Scraawl Account button at the bottom of the page.

5. How can I get started with a PixL account. Is there a free option?

If you would like to try PixL, please get in touch through our Contact page. There is no free option, but we do offer PixL demos for those interested.

6. What image and video formats are supported by PixL?

Almost all standard video and Image file types are supported in PixL. A complete list of image and video file formats supported by PixL is provided below:

Image file types: jpg, png, tga, raw, bmp, pcx, svg, webp, gif, tif, dib, ico, cr2, crw, dcr, dng, kdc, mef, mrw, nef, orf, pef, raf, sr2, srw, x3f, ppm, pgm, tiff, epi, eps, jpc, psd, pnm, rgb, rgba, sgi

Video file types: mp4, avi, rm, flv, wmv, 3gp, webm, rmvb, mpg, mts, m4v, mov, swf, ts, ogv, 3g2, f4v, h264, mxf, qt

7. Is there a maximum file size that PixL can handle?

The maximum files size is set by user policy and depends on the plan you have purchased. It can vary between 100MB to 4GB , depending on your policy. Contact customer support for more information.

8. How can I upload a video or image?

Click on Create Report. A new tab will open with the options to upload images or a video. Give your report a name, then click on the green Add File(s)/Search for File button. From here a window will pop up where you can select your video or image source. Now click on Upload Files and start Scraawling!

9. How do I know if a file has been uploaded?

If a file has successfully been uploaded, the upload tab will refresh and the text Upload Complete will appear on your screen. Do not close the upload tab while the video is uploading.

10. What happens to the videos or images once they are uploaded to PixL?

Once files are uploaded they will remain accessible in your My Reports page until you delete them.

11. Do I need any machine learning expertise to use PixL?

No machine learning experience is needed to use PixL. PixL image and video analytics algorithms have already been trained with large amounts of ground truth data. Users can simply invoke these analytics on their data from the PixL user interface and view the results.

1. How many reports can I set up and how many files can I upload per month?

The number of reports allowed and files per month depends on your account privileges. Go to the My Profile page to see your account privileges.

2. I get a message that I have to delete a report before creating a new one, why?

Your license allows you to have a certain number of reports. You will get this message if you’ve reached your limit and are trying to create a new report. You can delete one or more of your existing reports to get around this. Different licenses have different report limits. Contact us if you are interested in upgrading your account.

1. How do I zoom or pan?

In order to pan, you must first zoom in on the media. To zoom in on a video, click on the magnifying glass symbol in the options bar at the bottom of the player. Once you’ve zoomed in, you can use your left and right keys to pan left and right. To zoom in on an image, select the plus symbol from the options bar. Use your mouse to adjust the image until the desired area is in view.

2. How do I speed up the video?

At the bottom of the video you should see the symbol 1x. If you hover over the 1x you will see options for video speed.

3. Can I reduce the brightness of a video?

You can edit brightness, contrast, saturation, and hue of any video by hovering on the hamburger menu in the options bar.

4. Are the changes I make in the player permanent?

Not at all, you can reset your view, speed, and other edits by clicking on the original setting or clicking on Reset from the hamburger or miscellaneous menus.

5. The player won’t let me change to Face Detection View. Why is that?

This is because you have not yet run the analytic. Once you do, then you will be able to view the detected faces in the player.

6. What details does Metadata discover?

The Metadata tab in the player can reveal further details about a file including image type, image size, date of last modification, embedded location data, etc.

1. What analytics are included in PixL?

Pixl includes Face Detection, Face Recognition, Object Detection and classification, face and object tracking, Co-occurrence analysis, and Custom Detection for Weapons, Extremist Flags, and Military Objects.

2. I clicked on the Details button on an analytic but there is nothing there.

To view an analytic, select Details and then click on Run Analytics in the yellow bar.

3. Why is a box greyed out?

A grey box indicates that an analytic is in the process of development or that in order to access the analytic you must upgrade your plan.

4. Why aren’t the analytics automatically run?

Some of the analytics offered by PixL are computationally heavy. For faster and more distributed load times, analytics must be run manually by users.

5. What are the factors that impact the quality of PixL’s detection and analytics results?

Several factors such as resolution, motion blur, pose, partial occlusions, and lighting conditions may impact the quality of the results.

6. How small can an object or face be for PixL detect and analyze it?

PixL is currently optimized to detect objects and faces that are a minimum of 40 pixel in either dimension.

7. What is the difference between Face Detection and Face Recognition?

Face Detection looks for scenes where a face occurs, whereas Face Recognition will match a face found in your report to a reference data set of known faces.

8. How can I find the other instances of a face?

Tracks of similar faces are grouped into a single entity. Users can drill down into analytics or the player to navigate through the detections and tracks.

9. What is Co-occurrence Analysis?

PixL’s Co-occurrence analytic identifies when and where the tracks of entities overlap in an image or video. For each overlap, or co-occurrence, details of the time of overlap, frame IDs, duration of overlap, number of overlaps, and details of overlapping faces are provided. This analytic provides the capability to identify two or more people who appear in the same scene and other scenes in which the same individuals are seen together. PixL’s interactive user interface allows you to search through co-occurrences and view the frames associated with each co-occurrence, giving context to the instance of each overlapping track.

10. What does PixL Face Recognition Analytics do?

PixL’s Face Recognition analytics allow you to recognize individuals of interest by matching faces detected within a video or set of images against a user-provided dataset of faces. A confidence level is returned with each match. Currently PixL returns 4 confidence levels – High, Medium, Low, and Very Low.

11. What is a recognition dataset and how do I create one?

A recognition dataset is a searchable repository of labeled faces, which is either created and owned by the user or provided by PixL, for example the celebrity dataset. The repository may contain more than one face of an individual. The faces in the recognition dataset are converted and saved as feature vectors, which are used to compare a detected face against that of a face in the recognition dataset.

12. Does PixL come with any pre-configured datasets?

PixL’s pre-configured datasets include Extremists, World Leaders, U.N. Sanction List, and Celebrities.

13. How do I create and edit a reference dataset?

To create a dataset, upload the images containing your reference faces. Navigate to Face Detection, select Faces, and click View All. From here you can select individual faces to add to your reference dataset by clicking the checkbox under each face, or add all of the faces by clicking Select All. Once you have made your selection, select the Action menu and click Export. You will have the choice to add faces to an existing dataset or a new dataset. Once the dataset is created users can edit and relabel datasets using the PixL recognition datasets viewer and editor.

14. About how many faces can PixL accurately recognize?

For commercial use cases, PixL comes pre-configured with a celebrity dataset for celebrity recognition. The celebrity dataset includes over half a million images of 50,000 celebrities. PixL provides users with the tools needed to update, label, and maintain their own custom image/face datasets against which face recognition can be performed.

15. How does PixL perform Object Detection?

PixL uses multiple advanced artificial intelligence and machine learning models to detect and label objects. PixL machine learning-based models predict the positions of bounding boxes, the object classes, and their probabilities directly using fully connected layers on top of a convolutional feature extractor. The Object Detection modules are combined with a robust tracker to track and predict the positions of the detected objects. Objects are classified into 80 classes/labels (e.g., car, truck, motorcycle, cat, dog, bird, etc.), and these labels are grouped and displayed in 4 high level categories – People, Vehicles, Animals, and Other.

16. What sort of objects does object tracking detect and follow?

Scraawl PixL can detect and track thousands of objects such as vehicles, buildings, and even household goods.

17. What are Custom Detectors?

PixL’s Custom Detectors can detect Extremist Flags, Military Weapons, or Military Equipment.

1. What does the Track Data menu option in PixL contain?

Track Data contains detailed information of all entities that were detected in the upload video or images, including the entity ID, the track ID, the number of overlaps, number of frames that the entity appeared in, and track start and end times. All the track data is searchable by entity ID’s and other parameters such as overlaps, track IDs, etc using Boolean search operators such as a AND, OR and NOT.

Terms of ServiceTerms of Service

Read our approved use cases and more
Back to top