ML Research Hub

0:25

🌟 DG-Mesh: Constructing high-quality polygonal meshes from monocular video.

DG-Mesh reconstructs a high-quality dynamic 3D vertex-matched mesh from monocular video. The pipeline uses 3D Gaussian wavelets to represent dynamic scenes and differentiable algorithms to construct polygons.

DG-Mesh allows you to track the movement of vertices, simplifying the texturing of dynamic objects.
The method is memory efficient and fully differentiable, allowing optimization of the target object's 3D mesh directly.

The Github repository contains code for local training using datasets:

- D-NeRF
- DG-Mesh
- NeuralActor
- Custom dataset , shot on Iphone 14 Pro and processed in Record3D, RealityCheck and masked in DEVA.

🖥 Local launch:

conda create -n dg-mesh python=3.9
conda activate dg-mesh
conda install pytorch torchvision torcaudio pytorch-cuda=11.8 -c pytorch -c nvidia

#Install nvdiffrast
pip install git+https://github.com/NVlabs/tiny-cuda-nn#subdirectory=bindings/torch
pip install git+https://github.com/NVlabs/nvdiffrast/

# Install pytorch3d
export FORCE_CUDA=1
conda install -c fvcore -c iopath -c conda-forge fvcore iopath -y
pip install "git+https://github.com/facebookresearch/pytorch3d.git"

# Clone this repository
git clone https://github.com/Isabella98Liu/DG-Mesh.git
cd DG-Mesh

# Install submodules
pip install dgmesh/submodules/diff-gaussian-rasterization
pip install dgmesh/submodules/simple-knn

# Install other dependencies
pip install -r requirements.txt

🟡

Project page

🖥

GitHub

🟡

Arxiv

#Video2Mesh #3D #ML #NeRF

https://t.iss.one/DataScienceT

⭐️

Please open Telegram to view this post

VIEW IN TELEGRAM

Please open Telegram to view this post

VIEW IN TELEGRAM

👍4❤1

3.02K views10:09

Forwarded from Advert Deal

📈Professional trader that post daily analysis and recommendations. His analytics and signals were stolen by almost all telegram channels.

You don't have to pay any fees, just join.

https://t.iss.one/+_kCQKqG5SNZkODhi

2.31K views14:04

📄Deep learning applications in single-cell genomics and transcriptomics data analysis

📘Journal: Biomedicine & Pharmacotherapy (I.F.=6.9)
🗓Publish year: 2023

🧑‍💻Authors: Nafiseh Erfanian, A. Ali Heydari, Adib Miraki Feriz,...
🏢University: Birjand University of Medical Sciences, Iran - University of California, Merced, USA - University of Calgary, Calgary, Canada, ...

📎 Study the paper

#review #deep_learning #single_cell #genomics #transcriptomics

❤2👍2

2.64K viewsedited 15:05

This infographic delves deeper into the difference between Open Source and Closed Source LLMs.

Become an #LLM master! Enroll in our exclusive 5-day LLM Bootcamp (online & in-person)

Link: https://datasciencedojo.com/bootcamps/large-language-models-bootcamp/

https://t.iss.one/DataScienceT

⭐️

Please open Telegram to view this post

VIEW IN TELEGRAM

👍4

2.49K viewsedited 17:42

0:11

Vector Database by Hand ✍️ Make Your Own 👉 by-hand.ai/s/vecdb

Previously I shared a Google Sheet to make custom AI by Hand ✍️ exercises for the Transformer. Thousands of people made copies of the spreadsheet. Thank you! 🙏

Encouraged, I am following up with a similar tool for Vector Database. I am trying my best to match the layout of the matrices in the original exercise I shared earlier.

To make your own custom version, simply follow the link above to create a copy of the spreadsheet. Try changing some weights, biases, words, and even the word embeddings. See how the calculation changes accordingly.

If you are teaching a course, you can hide the answers and print a copy to challenge your students! I promise this will make you really popular! 😉

https://t.iss.one/DataScienceT

⭐

Please open Telegram to view this post

VIEW IN TELEGRAM

👍14

3.04K viewsedited 04:02

🩵 940+ FPS Multi-Person Pose Estimation 💛

👉 RTMW (Real-Time Multi-person Whole-body pose estimation models) is a series of high-perf. models for 2D/3D body pose estimation. Over 940 FPS on #GPU! Code & models 💙

🟡 Review: https://t.ly/XkBmg

🟡 Paper: arxiv.org/pdf/2407.08634

🟡 Repo: github.com/open-mmlab/mmpose/tree/main/projects/rtmpose

https://t.iss.one/DataScienceT

🏆

Please open Telegram to view this post

VIEW IN TELEGRAM

👍3❤2❤‍🔥2

3.08K viewsedited 15:58

Support Vector Machine Notes 🗒️ .pdf

8.6 MB

Support Vector Machine Notes

#SVM #machineLearning #AI #python

https://t.iss.one/codeprogrammer

⭐️

🚓 Support us:

🥇

BTC:
bc1qgjmr3ffh48jw5vw2tqad9useumutt5tql0pa6w

💲

USDT TRC20:
TMzAr8AFcZ1n5RXZa3BHPXHBRqugx9Skr7

🐬

TON:
UQAVMaOmfh8vsaXTykpBX45A3tsYv4Guo09eMw1Tl_uSYFcq

🏦

PAYPAL:
https://www.paypal.me/HusseinSheikho

Please open Telegram to view this post

VIEW IN TELEGRAM

👍4

3.22K views05:41

Forwarded from Machine Learning with Python

☄️ Did you miss our in-person Gen AI Rewind event in Bangalore? No worries! We're bringing the AI insights directly to your screen.

🔍 What We'll Cover:

2024's GenAI Highlight:
A concise yet comprehensive overview of the major breakthroughs you need to know.

Building GenAI Intuition:
Understand the fundamentals and grasp how GenAI is evolving at breakneck speed.

Crystal Ball Time:
Insights from our panelists' predictions for the AI landscape in the latter half of 2024.

Actionable Insights:
Learn how to leverage GenAI in your field and stay ahead of the curve.

👥 Who Should Attend?

Whether you're an AI enthusiast, a curious professional, or someone looking to future-proof their career, this session is tailored for you. No advanced technical knowledge required – just bring your curiosity!

This event presents a unique opportunity to learn from AI experts, expand your knowledge, and network with fellow AI enthusiasts from various industries.

Join our community of innovators and stay at the forefront of AI technology

❤1

1.75K views15:34

Forwarded from Machine Learning with Python

🚀 New Telegram Channel Alert!
We've discovered a gem for AI enthusiasts: AI.News.Daily!

Why follow AI.News.Daily? It's your go-to for:

• Daily AI Breakthroughs: Get updates every day.
• Expert Insights: Learn from the pros with easy-to-understand advice.
• Monetization Tips: Turn AI into profit.

Top picks from the channel:

• ChatGPT Prompts: Boost your productivity with these 7 mind-blowing prompts.
• Website Building: Explore top AI tools for web development.
• Comprehensive Guide: Dive into a 45-page tutorial on AI models.
• Word Processing Tools: Discover the best AI tools for all your writing needs.
• AI Selection Tool: Compare AIs by price, view detailed graphs, and explore leaderboards.

Curious? There's more to explore!

👉Check this out: @ainews_daily

👍4

1.47K views17:00

🙃

Prediction of the winning country of the 2024 Olympics

👨🏻‍💻 Our project is designed to predict the winner of the 2024 Olympics and use historical data to train a machine learning model.

📄 In this project, we train a machine learning model based on historical data by using the number of medals of countries participating in the 2021 Tokyo Olympics. This dataset includes information such as the number of medals, demographic information of countries and economic indicators. Then, based on the predicted medals, we will make a ranking to determine the winning country.

🖥 From the dataset and coding to analysis and project results, all are available in the following GitHub repo.👇

┌ 💸 Predictive Olympic Winner 2024
├ 📃 Report
└ 🐱 GitHub-Repos

https://t.iss.one/codeprogrammer

⭐️

☕️

Coursera Membership Free

Please open Telegram to view this post

VIEW IN TELEGRAM

👍4❤‍🔥1😍1🏆1

2.52K views07:45

Forwarded from 🐳

0:27

🐳 @whale – #1 licensed platform gaming and sportsbook on Telegram!

2mil+ people trust us, 200к+ native users on @whalesocials, and the community is only growing!😈

❤️‍🔥Our buns ❤️‍🔥

🥰Supports BTC, USDT, TON, CELO and NOT
🤑Instant withdrawals
🥰Regular giveaways

Share your thoughts and feedback of @Whale on Ton.app and Trustpilot. You make us better ❔

Please open Telegram to view this post

VIEW IN TELEGRAM

👍2❤‍🔥1❤1

1.53K views11:17

🔥 Play now 🔥

💬

72 hours of free access
💻 To all courses on the 365DataScience site
🔻 Free access to data science courses
✅ Without any restrictions!
📆 From July 19 to July 22

👨🏻‍💻 365DataScience website has created free and unlimited access to data science courses for 72 hours from November 19 , without any restrictions and without the need for a credit card!💯

⬅️ From today! You have the opportunity to access all the following for free by creating an account with an email.

🔹

81 training courses
🔹 672 practical and applied exercises
🔹 Over 260 hours of instructional videos

🚀 By taking these courses, you can experience a completely organized roadmap, test your knowledge with practical exercises, work on real projects, and finally receive fully valid certificates at no cost.

🔸

Course access link: 👇

┌ 💸 Free Access July 19 – July 22
└ 🎬 +81 Data & AI Courses 100% FREE

https://t.iss.one/codeprogrammer

⭐️

☕️

Please open Telegram to view this post

VIEW IN TELEGRAM

👍6❤2

2.96K views06:37

Aligning Sight and Sound: Advanced Sound Source Localization Through Audio-Visual Alignment

🖥

Github: https://github.com/kaistmm/SSLalignment

📕

Paper: https://arxiv.org/abs/2407.13676v1

🚀 Dataset: https://paperswithcode.com/dataset/is3-interactive-synthetic-sound-source

https://t.iss.one/DataScienceT

⭐

Please open Telegram to view this post

VIEW IN TELEGRAM

👍2

2.67K viewsedited 12:52

0:04

🎭 TRG: new SOTA in 6DoF Head 🎭

👉 ECE (Korea) unveils TRG, a novel landmark-based method for estimating a 6DoF head pose which stands out for its explicit bidirectional interaction structure. Experiments on ARKitFace & BIWI confirm it's the new SOTA. Source Code & Models to be released 💙

🤣 Review: https://t.ly/lOIRA

🤣 Paper: https://lnkd.in/dCWEwNyF

😏 Code: https://lnkd.in/dzRrwKBD

https://t.iss.one/DataScienceT

⭐

Please open Telegram to view this post

VIEW IN TELEGRAM

👍6

2.72K viewsedited 06:36

📢 Zuckerberg Releases Free ChatGPT Competitor: Llama 3.1!

Mark Zuckerberg just launched Llama 3.1, a next-gen AI model with the largest dataset ever. Available in 8B, 70B, and 405B versions, it boasts a 128k token context size.

Key Highlights:

• Performance: Outperforms GPT-4o and Claude 3.5 in general knowledge, math, and translation.
• Accessibility: Downloadable by anyone, bringing advanced AI within reach.
• Strategic Move: Meta applies pressure on OpenAI with this open-source release. You probably now understand why OpenAI showed GPT-4o mini a week ago and made it so cheap - soon we will have very smart models that run very fast on any hardware.

Try It Now:

• on Hugging Face: Llama 3.1 on Hugging Face
• on NVIDIA's website: NVIDIA’s website

This release represents a major development in open-source AI, potentially allowing broader access to advanced language models.

https://t.iss.one/DataScienceT

⭐

Please open Telegram to view this post

VIEW IN TELEGRAM

Please open Telegram to view this post

VIEW IN TELEGRAM

❤3👍2

2.49K views20:06