Which attack allows the attacker to know the steganography algorithm as well as the original and stego-object and extract the hidden information with the information at hand?

Which attack allows the attacker to know the steganography algorithm as well as the original and stego-object and extract the hidden information with the information at hand?

• A. Stego-only Attack
• B. Chosen-message Attack
• C. Blind Classifier attack
• D. Stego-only Attack

Answer: (D)

The scenario is about recovering the hidden payload when the attacker has the embedding method, the original cover, and the stego object. When you know exactly how data was embedded (the algorithm) and you also have both the original cover and the resulting stego, you can compare them directly and reverse the embedding process to extract the hidden information. This is the idea behind a known-original attack: access to the original content plus knowledge of the embedding technique enables straightforward extraction of the hidden data. This fits poorly with stego-only attacks, which assume you only have the stego object and nothing about the original or the embedding method. It also doesn’t describe a chosen-message attack (where the attacker can influence the message to be embedded) or the blind classifier approach (which focuses on detecting steganography rather than recovering the hidden data).

The scenario is about recovering the hidden payload when the attacker has the embedding method, the original cover, and the stego object. When you know exactly how data was embedded (the algorithm) and you also have both the original cover and the resulting stego, you can compare them directly and reverse the embedding process to extract the hidden information. This is the idea behind a known-original attack: access to the original content plus knowledge of the embedding technique enables straightforward extraction of the hidden data.

This fits poorly with stego-only attacks, which assume you only have the stego object and nothing about the original or the embedding method. It also doesn’t describe a chosen-message attack (where the attacker can influence the message to be embedded) or the blind classifier approach (which focuses on detecting steganography rather than recovering the hidden data).