Harpreet Sahota
Host at The Artists of Data Science
🤖 👨🏽💻 Hacker-in-residence @voxel51| I ❤️open source deep learning
Articles
-
2 months ago |
medium.com | Harpreet Sahota
Seriously, Memes Are All We NeedHarpreet Sahota·FollowPublished inVoxel51·10 min read·--A dataset of memes in FiftyOne formatThe idea for this project occurred when the Janus Pro model by DeepSeek AI dropped. I was reading through the paper, and in my hyped frame of mind, I thought I read the model excels in MEME Perception…in fact, what the paper actually said was the model excels in MME Perception.
-
2 months ago |
medium.com | Harpreet Sahota
Exploring the Intersection of Vision Language Models and Audio DataHarpreet Sahota·FollowPublished inVoxel51·13 min read·--ESC-10 dataset parsed as spectrograms into a FiftyOne DatasetI recently came across a paper that made me wonder: could we actually use vision language models to understand audio? The paper, Vision Language Models Are Few-Shot Audio Spectrogram Classifiers, introduces and explores Visual Spectrogram Classification (VSC).
-
2 months ago |
medium.com | Harpreet Sahota
Exploring the World’s Largest Insect Dataset with a Modern Toolkit for Visual AIHarpreet Sahota·FollowPublished inVoxel51·9 min read·--BIOSCAN in FiftyOneA new, comprehensive dataset called BIOSCAN-5M was introduced to the machine learning community at NeurIPS 2024, and it is a wealth of multi-modal information on over 5 million arthropod specimens, 98% of which are insects. Look, I get it; I’m as creeped out by bugs as the next guy.
-
2 months ago |
medium.com | Harpreet Sahota
Move over, CLIP — you’ve been dethroned!Harpreet Sahota·FollowPublished inVoxel51·8 min read·--Source: AIMv2 technical blogReleased in late 2024, AIMv2 is a family of open-vision encoders that has quietly revolutionized multimodal learning yet has received surprisingly little fanfare given its capabilities. What is AIMv2? AIMv2 is a family of pre-trained vision encoders that uses a novel multimodal autoregressive method.
-
Dec 6, 2024 |
medium.com | Harpreet Sahota
Where Research Meets Real-World Data ChallengesHarpreet Sahota·FollowPublished inVoxel51·6 min read·--Despite practitioners universally acknowledging that data quality is the cornerstone of reliable AI systems, only 56 out of 4,543 papers at NeurIPS 2024 explicitly focused on data-centric AI approaches. While this represents a doubling from 2023’s 28 papers, it remains a surprisingly small fraction given data’s outsized role in real-world AI success.
Try JournoFinder For Free
Search and contact over 1M+ journalist profiles, browse 100M+ articles, and unlock powerful PR tools.
Start Your 7-Day Free Trial →X (formerly Twitter)
- Followers
- 7K
- Tweets
- 6K
- DMs Open
- Yes
RT @NielsRogge: Excited to share 2 new notebooks I worked on! I called them "How I use VLMs in 2025" as they showcase my workflow on getti…
You gotta give flowers to people while you can. This is an appreciation post for @NielsRogge. He doesn’t know it, but he’s taught me so much. Absolute legend. 💐
RT @gowthami_s: Talkers don’t ship, shippers don’t talk. Choose wisely where you wanna be on that spectrum! #bayareagyan
