Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
enoch2090
48 days ago
|
parent
|
context
|
favorite
| on:
Meta Segment Anything Model 3
Surprisingly, SAM3 works bad on engineering drawings while SAM2 kinda works, and VLMs like Qwen3-VL works as well
zubiaur
48 days ago
|
next
[–]
Had good luck with Gemini 2.5, SAM3 failed miserably with PIDs.
retinaros
48 days ago
|
prev
[–]
yeah I tried too. Im trying a fine tuning on PIDs.
enoch2090
47 days ago
|
parent
[–]
Looking forward to your progress! Just checked the paper and it says the underlying backbone is still DETR. My guess would be that SAM3 uses more video frames during the training process and caused the dilution of sparse engineering-paper-like data.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
