ZipIt! Merging Models from Different Tasks without Training

ZipIt allows to combine completely distinct models with different initializations, each solving a separate task, into one multi-task model without any additional training.

🖥 Github: https://github.com/gstoica27/zipit

⏩ Paper: https://arxiv.org/abs/2305.03053v1

📌 Dataset: https://paperswithcode.com/dataset/nabirds

https://t.iss.one/DataScienceT

🔈Text-to-Video: The Task, Challenges and the Current State

In this post, we covered the constraints, unique challenges and the current state of text-to-video generation models

🤗 Hugging face: https://huggingface.co/blog/text-to-video

🖥 Github: https://github.com/huggingface/blog/blob/main/text-to-video.md

⏩ Damo-vilab: https://huggingface.co/damo-vilab

📌 Dataset: https://m-bain.github.io/webvid-dataset/

https://t.iss.one/DataScienceT

Money doesn't grow on trees, but don't worry friend, it grows on our trading accounts! 🤑

Every day I get messages from my subscribers that they are quitting their jobs because of me. Isn't that the best thanks for my hard work? 

🚀 For a limited time only, the first 30 subscribers will have free access to our VIP channel. After that, there will be a fee to enter.

👇 Click on the link below to join our Telegram channel and start trading smarter today!
Join the channel now (CLICK)

🚘 I BOUGHT MYSELF A NEW CAR!

Many thanks to this 👉🏼 guy for giving me a great way to make money in these tough times!🔝

🙌 Now I sold my old car and bought a new one that I've been dreaming about for years. 
Dreams come true!

😎 You just have to seize the chance in time!

Anyone interested, here's a link to his channel 👇🏼
https://t.iss.one/joinchat/ckyJ98FGQY85MDI8

🔥 ImageBind: One Embedding Space To Bind Them All

ImageBind, an approach to learn a joint embedding across six different modalities - images, text, audio, depth, thermal, and IMU data.

🖥 Github: https://github.com/facebookresearch/imagebind

Ⓜ️ Meta blog: https://ai.facebook.com/blog/imagebind-six-modalities-binding-ai/

⏩ Paper: https://arxiv.org/pdf/2305.05665v1.pdf

⭐️ Demo: https://imagebind.metademolab.com/

📌 Dataset: https://paperswithcode.com/dataset/msr-vtt

https://t.iss.one/DataScienceT

Object Detection Using Mask R-CNN with TensorFlow 2.0 and Keras.

https://blog.paperspace.com/mask-r-cnn-tensorflow-2-0-keras/

Mask_RCNN project: https://github.com/matterport/Mask_RCNN

https://t.iss.one/DataScienceT

🔰 REST APIs with Flask and Python in 2023

🌟 4.6 - 20097 votes 💰 Original Price: $74.99

Build professional REST APIs with Python, Flask, Docker, Flask-Smorest, and Flask-SQLAlchemy

Taught By: Jose Salvatierra, Teclado by Jose Salvatierra

Available now in Paid channel

Price: 5$ + free subscription in paid channel

Payment method: PayPal, payeer, crypto - contact @Hussein_Sheikho

For visa card or MasterCard:
https://boosty.to/datascienceteam/donate

Link channel:
https://t.iss.one/+LnCmAFJO3tNmYjUy

📖 DaGAN++: Depth-Aware Generative Adversarial Network for Talking Head Video Generation

A novel self-supervised method for learning dense 3D facial geometry (ie, depth) from face videos, without requiring camera parameters and 3D geometry annotations in training.

🖥 Github: https://github.com/harlanhong/cvpr2022-dagan

⏩ Paper: https://arxiv.org/pdf/2305.06225v1.pdf

⭐️ Demo: https://huggingface.co/spaces/HarlanHong/DaGAN

📌 Dataset: https://paperswithcode.com/dataset/voxceleb1

https://t.iss.one/DataScienceT

A summary of any PDF file is now available thanks to the Transforms library, In addition, you can now ask a question to answer in the same file

🌟 By @CodeProgrammer

30% discount on subscription to the paid bot - only for the first ten subscribers

The new price is now $7 instead of $10

The bot includes more than 30 million data science books and other scientific fields, in addition to more than 80 million scientific articles

The bot has the advantage of sending books as PDF files and reading files.

The bot is easy to use, just enter the name of the book and you will get it directly as a PDF file

To subscribe to the bot: @Hussein_Sheikho

⭐️ Towards Building the Federated GPT: Federated Instruction Tuning

Shepherd: A lightweight, foundational framework enabling federated instruction tuning for large language models

🖥 Github: https://github.com/jayzhang42/federatedgpt-shepherd

⏩ Paper: https://arxiv.org/pdf/2305.05644.pdf

📌 Data Preparation: https://github.com/jayzhang42/federatedgpt-shepherd#Data_Preparation

https://t.iss.one/DataScienceT

Discover and Cure: Concept-aware Mitigation of Spurious Correlation

🖥 Github: https://github.com/wuyxin/disc

⏩ Paper: https://arxiv.org/pdf/2305.00650v1.pdf

💨 Dataset: https://paperswithcode.com/dataset/metashift

https://t.iss.one/DataScienceT

Pre-train and Search: Efficient Embedding Table Sharding with Pre-trained Neural Cost Models

🖥 Github: https://github.com/daochenzha/neuroshard

⏩ Paper: https://arxiv.org/pdf/2305.01868v1.pdf

https://t.iss.one/DataScienceT

Multimodal Data Augmentation for Image Captioning using Diffusion Models

🖥 Github: https://github.com/xiaochr/multimodal-augmentation-image-captioning

⏩ Paper: https://arxiv.org/pdf/2305.01855v1.pdf

💨 Dataset: https://paperswithcode.com/dataset/coco

https://t.iss.one/DataScienceT

⭐️ Towards Building the Federated GPT: Federated Instruction Tuning

Shepherd: A lightweight, foundational framework enabling federated instruction tuning for large language models

🖥 Github: https://github.com/jayzhang42/federatedgpt-shepherd

⏩ Paper: https://arxiv.org/pdf/2305.05644.pdf

📌 Data Preparation: https://github.com/jayzhang42/federatedgpt-shepherd#Data_Preparation

https://t.iss.one/DataScienceT

ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding

You can easily plug in any 3D backbone models and pre-train it using our framework to get a jump-start for various downstreaming tasks!

🖥 Github: https://github.com/salesforce/ulip

⏩ Paper: https://arxiv.org/abs/2305.08275v1

📌 Dataset: https://paperswithcode.com/dataset/objaverse

https://t.iss.one/DataScienceT

FastComposer: Tuning-Free Multi-Subject Image Generation with Localized Attention

FastComposer uses subject embeddings extracted by an image encoder to augment the generic text conditioning in diffusion models, enabling personalized image generation based on subject images and textual instructions with only forward passes.

🖥 Github: https://github.com/mit-han-lab/fastcomposer

⏩ Paper: https://arxiv.org/abs/2305.10431v1

📌 Dataset: https://paperswithcode.com/dataset/ffhq

⭐️ Project: https://fastcomposer.mit.edu/

https://t.iss.one/DataScienceT

Hybrid and Collaborative Passage Reranking

🖥 Github: https://github.com/zmzhang2000/hybrank

⏩ Paper: https://arxiv.org/pdf/2305.09313v1.pdf

💨 Dataset: https://paperswithcode.com/dataset/natural-questions

https://t.iss.one/DataScienceT

FunASR: A Fundamental End-to-End Speech Recognition Toolkit

FunASR, an open-source speech recognition toolkit designed to bridge the gap between academic research and industrial applications

🖥 Github: https://github.com/alibaba-damo-academy/FunASR

⭐️ Docs: https://alibaba-damo-academy.github.io/FunASR/en/index.html

⏩ Paper: https://arxiv.org/abs/2305.11013v1

📌 Dataset: https://paperswithcode.com/dataset/wenetspeech

https://t.iss.one/DataScienceT

Segment Any Anomaly without Training via Hybrid Prompt Regularization

This project addresses zero-shot anomaly detection by combining SAM and Grouding DINO.

🖥 Github: https://github.com/caoyunkang/segment-any-anomaly

🖥 Colab: https://colab.research.google.com/drive/1Rwio_KfziuLp79Qh_ugum64Hjnq4ZwsE?usp=sharing

⏩ Paper: https://arxiv.org/abs/2305.11013v1

📌 Dataset: https://paperswithcode.com/dataset/visa

https://t.iss.one/DataScienceT