Martin Wunderwald retweeted
Deepseek is AI that costs pennies.
Unitree is Robots that cost less than a MacBook. DeepSeek + Unitree G1 = the future of cheap, scalable digital labor. It’s not science fiction anymore. ❤️🤖
Readers added context they thought people might want to know
According to the Unitree website, pricing for the G1 starts at $16,000.00. This is significantly higher than the price of a MacBook.
unitree.com/mobile/g1
macprices.net/16-macbook-pro/
apple.com/shop/buy-mac/m…
Martin Wunderwald retweeted
btw, that parking lot management system I showed you yesterday can be built without any coding using @roboflow workflows.
Martin Wunderwald retweeted
Bird Buddy introduced the Smart Bird Feeder.
It lets you experience nature with an AI-powered camera feeder that captures photos of local birds and notifies you of visitors.
@mybirdbuddy features a high-resolution camera, recyclable plastic housing, and weather-resistant design.
You can receive photo postcards on your phone and get real-time alerts.
Martin Wunderwald retweeted
ManipGen can solve long-horizon tasks in the real-world entirely zero-shot generalizing across objects, poses, environments and scene configurations! We outperform SOTA approaches such as SayCan, OpenVLA, LLMTrajGen and VoxPoser across 50 tasks by 36%, 76%, 62% and 60%!
8/N
Martin Wunderwald retweeted
Can my robot cook my food, rearrange my dresser, tidy my messy table and do so much more without ANY demos or real-world training data?
Introducing ManipGen: A generalist agent for manipulation that can solve long-horizon robotics tasks entirely zero shot, from text input!
1/N
Martin Wunderwald retweeted
We built a GPT-4o-powered cleaning robot.
- $250 for the robot arms
- 4 days to build
Open source is truly democratizing the field of robotics.
@KasparJanssen
Martin Wunderwald retweeted
FruitNeRF: A Unified Neural Radiance Field based Fruit Counting Framework
arxiv.org/abs/2408.06190
Project: meyerls.github.io/fruit_nerf…
Code: github.com/meyerls/FruitNeRF
Uses NeRFs to count any fruit type directly in 3D.
The code is a Nerfstudio extension!
Method ⬇️ 1 | 2
Martin Wunderwald retweeted
Artificial intelligence detects breast cancer 5 years before it develops
#MedEd #MedTwitter #SCIENCE #technology #oncology #Cancer #Diagnosis
Martin Wunderwald retweeted
supervision-0.22.0 is coming out today
one of the things we release is Mediapipe integration along with default visualizers for face and body pose keypoints
link: github.com/roboflow/supervis…
Martin Wunderwald retweeted
what papers are absolute must-haves if I want to learn vision language models? here's what I have on my list for now:
1. Efficient Estimation of Word Representations in Vector Space (2013)
2. Attention Is All You Need (2017)
3. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding (2018)
4. Language Models are Unsupervised Multitask Learners (2018)
5. T5: Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer (2019)
6. BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension (2019)
7. Language Models are Few-Shot Learners (2020)
8. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (2020)
9. CLIP: Connecting Text and Images (2021)
10. ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision (2021)
11. ALIGN: Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision (2021)
12. BEiT: BERT Pre-Training of Image Transformers (2021)
13. LoRA: Low-Rank Adaptation of Large Language Models (2021)
14. Flamingo: a Visual Language Model for Few-Shot Learning (2022)
15. BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation (2022)
16. LLaMA: Open and Efficient Foundation Language Models (2023)
Martin Wunderwald retweeted
prototype of Florance-2 video processing space
- generate a caption using the first frame
- use the caption to phrase grounding to generate detections
link to @huggingface space: huggingface.co/spaces/Skalsk…
Martin Wunderwald retweeted
I finetuned an LLM for spam detection.
It reuses gpt-2 weights and has 124M params.
Total times:
• finetuning took 59 seconds
• predictions took 0.02 seconds
Everything was done in @GoogleColab for free.
Since it's a tiny model, finetuning and inference is extremely fast.
Next up: finetuning an LLM to detect sentiment in SEC documents.
Martin Wunderwald retweeted
I'm working on my Florence-2 tutorial and decided to add this slide at the end to put this model in the context of other hot CV models.
did I get this right? let me know if I messed up something. I still have a few hours before I include that in the video.
Martin Wunderwald retweeted
OMG. I was waiting for this to happen. Florence-2 is running in browser.
@xenovacom can I run any fine-tuned model or only pre-trained ones?
Martin Wunderwald retweeted
Some help here: huggingface.co/spaces/Ali-C1…
And here: blog.eleuther.ai/transformer…
Martin Wunderwald retweeted
[SIGGRAPH '24] RTG-SLAM: Real-time 3D Reconstruction at Scale Using Gaussian Splatting
Paper (pdf link): gapszju.github.io/RTG-SLAM/s…
Project: gapszju.github.io/RTG-SLAM/
Code: github.com/MisEty/RTG-SLAM
Method ⬇️ 1 I 2
Martin Wunderwald retweeted
A cyclist who understands physics well.
nitter.gq/tannokasa3/status/1797…
