BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.romhack.io//romhack-camp-2026//speaker//C9YMMS
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-romhack-camp-2026-HBCFWE@cfp.romhack.io
DTSTART;TZID=CET:20261003T164000
DTEND;TZID=CET:20261003T172000
DESCRIPTION:Deepfake detection using 3D CNNs traditionally focuses on ident
 ifying synthetic facial manipulations in video. This work inverts that len
 s. I leverage temporal‑spatial 3D CNN features originally designed to sp
 ot fake videos and apply them to a completely different domain: OSINT‑ba
 sed profiling. By treating metadata as a volumetric signal across related 
 individuals\, the same convolutional filters that detect frame‑to‑fram
 e anomalies can reconstruct familial linkages from public or leaked databa
 ses. This methodology enhances traditional OSINT by adding a predictive la
 yer. Instead of manually building family trees\, the 3D CNN model learns k
 inship patterns and flags potential person matches with high confidence. T
 he result is a hybrid attack that combines AI forensics and open source in
 telligence to identify a target individual from a distant relative’s spi
 t sample. I demonstrate a proof of concept using synthetic profiles and re
 al OSINT sources. The talk shows how a deepfake detection technique become
 s a privacy‑breaking weapon for person traceability at scale.
DTSTAMP:20260921T191304Z
LOCATION:STAGE 2 (Ghost in the Shellcode)
SUMMARY:Using Temporal‑Spatial 3D CNN Features to Enhance OSINT‑Based P
 rofiling and Individual Traceability - Reza
URL:https://cfp.romhack.io/romhack-camp-2026/talk/HBCFWE/
END:VEVENT
END:VCALENDAR