Introduction
Investigators need to sift through huge amounts of digital data in order to find the evidence they are looking for. In many cases, this evidence is an image or video. To make life easier for them, we have created the “Media category wizard”. They can create a new media category and by harnessing the power of machine learning, they can train Cellebrite’s Analytics to automatically find matches for them. The more examples they give the product, the better it gets at recognizing the category.
My role: UI designer | Team members: Yair Golan - UX designer, Aliza Solomon - UX writer | Tools used: Sketch, Illustrator, Photoshop. | This project was done as part of my work at Cellebrite.
My role: UI designer | Team members: Yair Golan - UX designer, Aliza Solomon - UX writer | Tools used: Sketch, Illustrator, Photoshop. | This project was done as part of my work at Cellebrite.
The challenge
- Automatically filing the media files into categories, so you can focus on what you're looking for.
- Saving precious time and get the results you want faster.
Pain points
- It’s frustrating going through lots of media files, in order to find the specific file you need.
- This process can take a long time.
- You might miss critical evidence due to too many files or time pressure issues.
Gain points
- By letting the product automatically do the work for you, you will save precious time.
- You can focus on the files that are relevant to the investigation and not go through all the clutter.
User flow
Personas
The user is an investigator who wants to find the relevant evidence and solve the crime.
Design
Scenario
In his investigation, Steve needs to find images of the murder weapon. He creates a new media category to help him track all the relevant images.
Create a new category
Steve opens the Media category wizard. He chooses a name for the category and adds a description.
Steve opens the Media category wizard. He chooses a name for the category and adds a description.
Sample images
Steve selects the images that match the category. He can view these images in the right panel.
Steve selects the images that match the category. He can view these images in the right panel.
Apply filters
Steve can filter the images by dates or tags. He can also upload new images at any time by clicking the blue floating "+" button.
Steve can filter the images by dates or tags. He can also upload new images at any time by clicking the blue floating "+" button.
Crop images
Steve can enlarge an image to view it. He can zoom in, zoom out, rotate and crop the image.
Steve can enlarge an image to view it. He can zoom in, zoom out, rotate and crop the image.
Review and confirm
The wizard presents images that it thinks match the category. Steve approves the images that are correct. The more images he reviews, the better the wizard gets at recognizing the category.
The wizard presents images that it thinks match the category. Steve approves the images that are correct. The more images he reviews, the better the wizard gets at recognizing the category.
Final thoughts
Working on this project was very fascinating, from learning about cutting-edge technology, to designing a simple to use, 3 step wizard. The impact this project has on the investigator’s work is enormous. We used machine learning to teach the product to identify images that are relevant to the investigation. This is a big time saver for investigators, and allows them to focus on other aspects of the investigation.