data
Video-Swin-Transformer
Our great sponsors
data | Video-Swin-Transformer | |
---|---|---|
116 | 7 | |
16,631 | 1,309 | |
0.4% | 0.5% | |
8.5 | 0.0 | |
about 2 months ago | about 1 year ago | |
Jupyter Notebook | Python | |
Creative Commons Attribution 4.0 | Apache License 2.0 |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
data
-
[USMNT] It only took 20 caps for Jesus Ferreira to get double-digit goals. The fastest in #USMNT history.
You of course already know this answer, but just to put it into more perspective. Here are the SPI ranking equivalents to what he did with these 11 goals in Scotland and Switzerland.
-
[Effortpost] Advanced stats on which players are contributing the most to the Heat's playoff run.
To answer these questions I decided to look at 538’s RAPTOR ratings. RAPTOR uses player tracking data to estimate how much each player contributes on the offensive and defensive ends. The total RAPTOR score should be something like the “number of points a player contributes to his team’s offense and defense per 100 possessions, relative to a league-average player.” Higher is better, best during the regular season has been Nikola Jokic at +14. You can read more about it here or play with an interactive tool on their website here. I don’t really care about the details of why it’s a good statistic, but it seems pretty helpful and most importantly for my purposes you can download the data here for free.
-
Consanguineous marriage percentage per country
EDIT: I came to this data from this repository which has a nice csv collection for machine training.
-
USMNT is a European club. How did they do this season?
Looks like we may actually be collectively underrating our guys now. That's an interesting change. Based on SPI (rating = 72.4) we would be:
- Derrick White's WAR over the past season has been ~6.7 according to a composite of various metrics. Derrick White's WAR in the playoffs has been ~0.1 according to RAPTOR. The worst among the main Boston roster
-
Nate Silver: Some personal news
Before Disney/ABC get any -ideas-, might be a good chance to get our hands on at least their data[0]!
[0]: https://data.fivethirtyeight.com/
-
In honor of Sexual Assault Awareness Month, make sure neither you nor friends harbor any misconceptions about consent
Most young women expect words to be involved when their partner seeks their consent. 43% of young men actually ask for verbal confirmation of consent. Overall, verbal indicators of consent or nonconsent are more common than nonverbal indicators. More open communication also increases the likelihood of orgasm for women.
- CMV: When selecting a movie to watch, the audience's rating is the only thing that matters and the critic's rating is entirely irrelevant.
-
Slight majority of people in WA want to leave state, poll finds
DHM does not use an equity sample. Of all polling operations they rank 250 out of 517. Id like to see another pollster https://github.com/fivethirtyeight/data/blob/master/pollster-ratings/pollster-ratings.csv
-
Optimism is bad for your health. So lets just do some maths! How can Liverpool FC get top 4? part 2
LOL My github’s pretty sparse but I’m pulling data from this API; 538 also provides the data they use for their club predictions here if that interests you
Video-Swin-Transformer
- Explanation needed
- Explanation needed [P]
- Explanation needed [R]
-
Weekly Entering & Transitioning Thread | 20 Feb 2022 - 27 Feb 2022
PROBLEM STATEMENT Develop an efficient common strategy and relevant implementation to extract the video-based models in the black box and grey box setting across the following 2 problem statements. 1.Action Classification Model Extraction for Swin-T Model for Action Classification on Kinetics-400 dataset. Download the model from here- https://github.com/SwinTransformer/Video-Swin-Transformer 2.Video Classification Model Extraction for MoViNet-A2-Base Model for Video Classification on Kinetics- 600 dataset Download the model from here- https://tfhub.dev/tensorflow/movinet/a2/base/kinetics-600/classification/3 Blackbox Setting Do not use any relevant data set available and use synthetic or generated data without using the Kinetics series dataset. Also, do not use the same model architecture as the original model to train the extracted model. Greybox Setting You can use 5% of original data (balanced representation of classes) as a starting point to generate the attack dataset. Also, do not use the same model architecture as the original model to train the extracted model. Can someone explain the problem statement in a easy / understandable way ?? What I think is the models have already been provided and we have to do something in Blackbox and greybox . Can someone explain in brief what we have to do in the blackbox / greybox??
-
Action recognition models for images
There are two main variants for the SWIN transformer the original SWIN transformer, official implementation here, and the Video SWIN transformer, official implementation here. Both architectures are very similar with the differences being mainly in the size of the input. The SWIN transformer pretrained on imagenet can be used as the backbone for different applications either image or video-based. In fact, the authors pretrained the original SWIN transformer on imagenet then they modified the input size and then fine-tuned it on video action recognition datasets. In your case, you can use the original SWIN transformer pretrained on imagenet then fine-tune it on your own dataset without modifying anything about the input size, since it is designed to process images.
-
[R] New Study Proposes CW Networks: Greater Expressive Power Than GNNs
The code is available on project GitHub. The paper Video Swin Transformer is on arXiv.
- [R] Video Swin Transformer: SOTA on Video Recognition (84.9% top 1 on Kinetics-400 and 69.6% top 1 on Something-Something V2)
What are some alternatives?
uawardata - The data behind uawardata.com
Swin-Transformer-Object-Detection - This is an official implementation for "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows" on Object Detection and Instance Segmentation.
tidytuesday - Official repo for the #tidytuesday project
Swin-Transformer-Tensorflow - Unofficial implementation of "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows" (https://arxiv.org/abs/2103.14030)
ydata-quality - Data Quality assessment with one line of code
MoViNet-pytorch - MoViNets PyTorch implementation: Mobile Video Networks for Efficient Video Recognition;
quilt - Quilt is a data mesh for connecting people with actionable data
Swin-Transformer-Semantic-Segmentation - This is an official implementation for "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows" on Semantic Segmentation.
CodeSearchNet - Datasets, tools, and benchmarks for representation learning of code.
Swin-Transformer - This is an official implementation for "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows".
datagen - Generates customer, sales reps, sales mgrs, products, manufacturer, and transaction data and creates and populates MySQL database with it. Also, can generate single tables of random data.
PaddleClas - A treasure chest for visual classification and recognition powered by PaddlePaddle