You're fine-tuning SD, right? You're aware that Samivel is in the original datasets too, yes? Why are you trying to avoid Copyrighted images on the fine-tune only? Might as well own it…
GENERATIVE AI
-
Companies continue AI recruitment focus despite market challenges
By
–
Yes Micha @MichaGUERIN
, c’est ce que j’ai écrit dans le #FlashTweet : ils se focalisent sur le recrutement. Trop tard ? Je ne sais pas, cela montre en tous les cas que le sujet n’est pas mort et enterré en tous les cas. Les entreprises continuent à itérer. C’est vide mais -
AI Model Training Data Requirements and Generalization
By
–
> No Samivel data was used in the training. I can't tell if you're joking… What are the chances of it "just knowing" without training data?
-
Daily tech roundup: AI models, Apple cash, Hopium losses
By
–
Les sources de l'édition du jour du #FlashTweet https://
ft.com/content/8bfaa5
00-feee-477b-bea3-84d0ff82a0de
… @FT https://
techcrunch.com/2023/11/02/sta
bility-ais-latest-tool-uses-ai-to-generate-3d-models/
… @TechCrunch https://
maddyness.com/2023/11/02/hop
ium-pertes-premier-semestre/
… @byMaddyness https://
cnbc.com/2023/11/02/app
le-now-has-162-billion-in-cash-on-hand-less-than-last-quarter.html
… @CNBC https://
coins.fr/assureur-swiss
-life-web3-avec-metaverse/
… @Coins_fr -
Tech giants embrace AI audits, Apple struggles, Swiss Life enters Metaverse
By
–
Édition du 3 novembre 2023 Chaque matin, #FlashTweet, le petit journal des grandes innovations, par @EmmanuelleL9 Les géants de la Tech acceptent l’audit de leur IA, Apple à la peine, Swiss Life se jette dans le Metaverse & plus Repost pour soutenir le brief tech
-
AI Image Captioning Tool for Emotional Salient Features Analysis
By
–
We also offer a captioning method to convert images into sentences describing emotionally salient features related to face, body, pose, human interactions, activity and environment. Social scientists might be interested in our manual annotation interface
-
Emotional AI: Fast Slow Emotion Recognition Vision Models
By
–
Captioning interface
Arxiv: https://
arxiv.org/abs/2309.13136
Project: https://
rosielab.github.io/emotion-captio
ns/
… Emotions Fast and Slow Arxiv: https://
arxiv.org/abs/2310.19995
Project: https://
rosielab.github.io/Fast-and-Slow/ -
Multimodal AI Shows Promise for Emotion Recognition Beyond Zero-Shot Methods
By
–
Overall, these zero shot methods do not perform nearly as well humans nor models tuned on the training data from EMOTIC (Emotions in Context). But there appears to be promise in linking vision with language to account for both affective and cognitive empathy
-
CLIP and LLaVA Struggle with Contextual Image Interpretation
By
–
In this example, we believe CLIP sees the "surprised cat pose" and predicts doubt, surprise and fear, ignoring context. Oddly, LLaVA has also gone a bit far, inferring the person in this image is experiencing sadness because it's their last ski trip of the season.
-
CLIP vs Captioning+GPT: Vision Model Comparison on Emotions
By
–
We compare CLIP (vision only) to Captioning+GPT (vision + reasoning) over 4714 images from the Emotions in Context dataset and observe a small but noticeable difference.