LLMs can see and hear without any training

https://github.com/facebookresearch/MILS

#HackerNews #LLMs #AI #MachineLearning #FacebookResearch #MILS #Innovation

Facebook က အစတည်းက အကျင့်သိက္ခာမရှိတဲ့ကုမ္ပဏီဆိုတာသိပေမယ့် ဒါကတော့ လွန်သွားပြီ။

https://techcrunch.com/2024/03/26/facebook-secret-project-snooped-snapchat-user-traffic/

#Facebook #Snapchat #FacebookResearch #privacy

Prompt Engineering with Llama 2

https://github.com/facebookresearch/llama-recipes/blob/main/examples/Prompt_Engineering_with_Llama_2.ipynb

"Prompt engineering is using natural language to produce a desired response from a large language model (LLM). This interactive guide covers prompt engineering & best practices with Llama 2." -- #facebookResearch

#gen_ai #llm #ml

🌖 GitHub - facebookresearch/audio2photoreal：從音頻生成逼真的照片級「Codec Avatars」的代碼和數據集
➤ 從音頻生成逼真的照片級「Codec Avatars」的代碼和數據集
✤ https://github.com/facebookresearch/audio2photoreal
這個項目包含了一個從音頻生成具有逼真照片級的「Codec Avatars」的代碼和數據集。該項目提供了訓練代碼、測試代碼、預訓練的運動模型以及數據集的訪問。如果您使用了數據集或代碼，請引用他們的論文。
+ 這項技術很有趣，可以用來創造逼真的虛擬影像。
+ 這個項目的代碼非常有用，可以幫助我們更好地理解和應用音頻和圖像之間的關係。
#GitHub #facebookresearch #audio2photoreal #Code and dataset #photorealistic Codec Avatars

Using segment-anything from #facebookresearch and
@MakiePlots
to create a little interactive segmentation UI for anything Makie can plot... E.g. zoomable satellite data with https://github.com/MakieOrg/Tyler.jl
The code (prototype status):
https://github.com/SimonDanisch/MakieSAM.jl

New Facebook AI fools facial recognition - The technology - which Facebook won't use in its own apps - subtly distorts face images so they're... more: https://nakedsecurity.sophos.com/2019/10/29/new-facebook-ai-fools-facial-recognition/ #artificialintelligence #de-identification #facialrecognition #facebookresearch #machinelearning #audioandvideo #biometrics #deepfakes #facebook #fakenews #privacy #ai