Truepic BlogGlossary: The ABCs of Authentic Digital Content

As the digital content landscape rapidly evolves, so does our lexicon. We put together a helpful glossary to help navigate the new world of synthetic content, generative AI, and digital verification. Let us know if you have a question or recommendation for more terms.

Asset Metadata

Asset metadata is additional information about an image or video such as time, date, camera model, file size and type, and author. It is viewable through the file property on any type of device (PC, Mac), as well as through Adobe programs (such as Photoshop and InDesign). It can be processed by machines, which allows it to be indexed, searched, verified for authenticity, and automatically processed.

Asset metadata can be manipulated through several accessible online programs, such as Photoscape, Daminion, AnalogExif, just to name a few. While sometimes it can be necessary to edit metadata for professional reasons, it can also be used to falsify images and videos.

In order to protect your business and customers from the threat of fraud, it’s important to verify that the images and videos in front of you are certifiably unaltered in any way, shape, or form. Just because an image may look authentic, it doesn’t mean that the metadata hasn’t been altered to tell a different story.

Cheapfake

A cheapfake is like a deepfake, but done through less expensive, non-artificially intelligent technology, such as Photoshop or video editing software. Cheapfakes attempt to deceive people through more conventional methods such as photo editing, app filters, recontextualization of media that already exists, selective editing, or even just altering the speed/pitch at which the media is replayed.

Cheapfakes are far more common and accessible to create as opposed to deepfakes. They can be used to forge signatures, create falsely incriminating images/videos, and distort the perception of information online.

While cheapfakes may be prolific across the internet, advancing security technologies have made it easier to ensure that timestamp, geolocation, internet comparison and more are verified as authentic.

Computer Vision

Computer vision is a field of artificial intelligence (AI) that enables computers and systems to derive meaningful information from digital images and videos. This then allows the AI to make recommendations based on the information it has collected. Another way of thinking about it is that computer vision allows a system to ‘see’ and make decisions from what it’s observed.

What makes computer vision unique is its ability to process more information, much more quickly, than a human would be able to. By utilizing deep learning, a type of machine learning, a system can ingest a large amount of data and teach itself how to differentiate each image from one another.

This can aid with everything from detecting defective products to verifying that uploaded images or video have not been manipulated. Computer vision can securely streamline tasks that would otherwise take a monumental amount of human effort, allowing businesses to focus on other key areas of their operation.

Control Capture™

Control Capture™ is Truepic’s breakthrough technology that ensures the integrity of digital photos and videos from point of capture. By using cutting-edge proprietary technology, it secures the capture process so that pixel contents and metadata, including date, time, and location, faithfully represent the what, when, and where of the visual reality before the camera.

Controlled capture technology is important for ensuring the authenticity and provenance of digital content, particularly in situations where the media is being used for legal or official purposes.

It uses sensors and other technologies, such as cryptography, to capture media in a controlled environment and stores the media on a local or cloud-based storage system for later review. This helps prevent tampering of the media and verifies that it’s a reliable record of the events or activities being captured.

Deepfake

Deepfakes are intricately manipulated images, videos or audio of people created through artificial intelligence (AI). The term itself comes from the AI technology that powers them: deep learning. This particular form of deep learning has taught itself to replicate the world around us, especially the human likeness. This ranges from swapping faces across images and videos, as well as mimicking audio. It can also generate realistic-looking photographs of people who have never existed, as well as catschemical compounds, and other everyday objects.

With the assistance of Generative Adversarial Networks (GANs), another popular method for creating deepfakes, flaws found in the initial deepfake can be detected and fixed, further making the image/video more convincing and harder to decode as fraudulent through technology. This can lead to scams such as fraudulent insurance claims for injury, lost objects, natural disasters, and much more.

As deepfakes become more sophisticated, getting ahead of the curve with digital content authenticity technology can help ensure that what you’re looking at is truly reality.

Machine Learning

Machine learning is a type of artificial intelligence (AI) that enables computers to learn and improve their performance of a task without being explicitly programmed by humans. It involves training a computer model on a large dataset and allowing the model to make predictions or decisions based on the patterns it finds in the data.

Machine learning helps by enabling computers to automatically learn from the large amounts of inputted data, and that includes sorting it, analyzing it, and identifying patterns of malicious activity. This all works together to help keep an organization’s information secure against malicious threats.

Optical Character Recognition

Optical character recognition (OCR), sometimes called text recognition, is the technology that allows the camera to read text on scanned documents or pictures taken, then display that text in a file to read, edit or search. OCR software analyzes the image and determines the shape, size and other characteristics of the text and compares them to a set of predefined templates or patterns.

By utilizing machine learning, image processing and pattern recognition, it can be used for a variety of applications, such as document management, data entry and automated billing systems. It’s also a great tool for organizations that need to digitize large volumes of text, and then search or edit the text quickly, saving time for all parties involved.

OCR can also aid in addressing misinformation, especially within the digital landscape. By extracting text from images or PDF files that contain false or misleading information, it can then run the text through a fact-checking database or algorithm to verify its accuracy.

Additionally, OCR can be trained to recognize and surface false or misleading information in text-based content. These use cases are especially important as cybercrime continues to change alongside emerging technologies.

Personally Identifiable Information

Personally identifiable information (PII) is any information that can be used to identify an individual. PII encompasses information such as name, address, date of birth, social security number, IP address, phone numbers, email addresses, and even financial or medical records. The main use of PII is by organizations who collect and store this data for business, financial, healthcare, and marketing research purposes.

The protection of PII is important because it can be used to commit identity theft or fraud if it falls into the wrong hands. Organizations that collect and store PII are often required to implement cybersecurity compliant measures to protect it from unauthorized access or disclosure. This can include measures such as encryption, access controls, and data breach notification policies.

Synthetic Media

Synthetic media is a field of digital content that has been created or modified by artificial intelligence (AI). This encompasses everything from text, images, audio, and video, and can cover pillars such as personalized media, generative AI, and deepfakes. It specializes in creating realistic simulations of people or events that have never actually happened or can be used to modify media in ways that weren’t previously possible via traditional methods.

While containing the potential for creative enhancement in areas such as education, entertainment or marketing, synthetic media can also be used to manipulate perceptions, erode trust of the media, and falsify images.

As AI continues to develop, it’s more important than ever to verify that the digital content you’re viewing is authentic and original.

Request more information