Nice video. If you want to replace the sky, you should select all of the sky and and few pixels into the sea and the island. Otherwise the ai assumes you want to keep the blue sky on the edges…
Great video. I’ve been wanting to try this and your walkthrough was a nice intro showing potential issues as well as how powerful this can be. I’d like to see you explore more of the deepfake audio and video stuff you discussed at the end as that could serve potentially useful in a Red Team engagement
Thanks mate! I've been seeing a lot about deepfake audio particularly recently, will definitely need to have a play around with the tooling at some stage. Interestingly, I was at an AppSecEU talk in like 2016-17, before making YT videos. The presentation was about voice-based passwords and how easy it will be to fake them, and how inherently flawed they are, since they can never be changed (unlike a regular password). They also talked about how content creators, streamers etc will be particularly vulnerable as there will be lots of public audio for attackers to make use of. So yeh, I guess I'm screwed 😂
![Leaking Secret Data with a Heap Overflow - "Leek" Pwn Challenge [Angstrom CTF 2023]](/img/n.gif)
Nice video. If you want to replace the sky, you should select all of the sky and and few pixels into the sea and the island. Otherwise the ai assumes you want to keep the blue sky on the edges…
Thanks 🙂 I noticed that overlapping by a few pixels when expanding the image produces better results, so your suggestion makes sense!
Great video. I’ve been wanting to try this and your walkthrough was a nice intro showing potential issues as well as how powerful this can be. I’d like to see you explore more of the deepfake audio and video stuff you discussed at the end as that could serve potentially useful in a Red Team engagement
Thanks mate! I've been seeing a lot about deepfake audio particularly recently, will definitely need to have a play around with the tooling at some stage. Interestingly, I was at an AppSecEU talk in like 2016-17, before making YT videos. The presentation was about voice-based passwords and how easy it will be to fake them, and how inherently flawed they are, since they can never be changed (unlike a regular password). They also talked about how content creators, streamers etc will be particularly vulnerable as there will be lots of public audio for attackers to make use of. So yeh, I guess I'm screwed 😂
For the fish in the water PS expert Jesus Ramirez has a great new video as adjusting intensity
Just watched it. Very cool! Thanks ☺
🫶🫶🫶
💜💜💜