The goal of the Kinetics dataset is to help the computer vision and machine learning communities advance models for video understanding. Given this large human action classification dataset, it may be possible to learn powerful video representations that transfer to different video tasks.
The Kinetics-700-2020 dataset will be used for this challenge. Kinetics-700-2020 is a large-scale, high-quality dataset of YouTube video URLs which include a diverse range of human focused actions. The aim of the Kinetics dataset is to help the machine learning community create more advanced models for video understanding. It is an approximate super-set of both Kinetics-400, released in 2017, Kinetics-600, released in 2018 and Kinetics-700, released in 2019.
The dataset consists of approximately 650,000 video clips, and covers 700 human action classes with at least 700 video clips for each action class. Each clip lasts around 10 seconds and is labeled with a single class. All of the clips have been through multiple rounds of human annotation, and each is taken from a unique YouTube video. The actions cover a broad range of classes including human-object interactions such as playing instruments, as well as human-human interactions such as shaking hands and hugging.
More information about how to download the Kinetics dataset is available here.
DevExpress 18.1 supports Windows 7 SP1 or later, Windows 8.1, Windows 10, Windows 11, and Windows Server 2012 R2 or later.
| Requirement | Details | |--------------|---------| | .NET Framework | 4.0 – 4.7.1 | | Visual Studio | 2010 – 2017 (also 2019 with some compatibility) | | OS | Windows 7 SP1 or newer (Server 2008 R2+) | | SQL Server | 2008+ (for reporting & data sources) |
: Go to DevExpress > Tools > Project Converter v18.1 inside Visual Studio to automatically update assembly references.
Contact support. Your account permissions may need to be refreshed for legacy downloads. Devexpress 18.1 Download
Registered users can securely access installer files directly via the DevExpress Download Manager .
Article last updated: 2025. All trademarks property of their respective owners.
If trying to use v18.1 with Visual Studio 2019, you might find that the components do not appear in the toolbox, even if the application compiles. DevExpress 18
: Windows 7, Windows 8, Windows 10, and Windows Server equivalents (both 32-bit and 64-bit).
To download DevExpress 18.1, users should use official channels to ensure security and license compliance: New in v18.1 - DevExpress
Downloading (released around May 2018) is often necessary for maintaining legacy enterprise applications, particularly those built with older .NET frameworks or older Visual Studio versions. Because it is an older version, it is not listed on the main download page and requires access to the DevExpress Download Manager using valid account credentials. Your account permissions may need to be refreshed
A DevExpress license key is (e.g., v18.1, v19.1, etc.). When you upgrade to a new major version, you must download an updated license key. However, a license key for a newer version is compatible with previous versions—meaning a v25.x license will also work with v18.1.
Open the from your Windows Start Menu. Run the repair process. Reopen Visual Studio as an Administrator. Registration and Licensing
Given the age of DevExpress 18.1 and the rapid evolution of the tech landscape, there are critical factors to consider for security and compliance.
Thus, the need for a reliable is not about “being outdated”—it’s about practical continuity.
No. DevExpress products are commercial software. You must purchase a license or use the 30‑day trial version of the current release. The free trial does not grant access to older versions. Legitimate access to v18.1 requires an active license.
1. Possible to use ImageNet checkpoints?
We allow finetuning from public ImageNet checkpoints for the supervised track -- but a link to the specific checkpoint should be provided with each submission.
2. Possible to use optical flow?
Flow can be used as long as not trained on external datasets, except if they are synthetic.
3. Can we train on test data without labels (e.g. transductive)?
No.
4. Can we use semantic class label information?
Yes, for the supervised track.
5. Will there be special tracks for methods using fewer FLOPs / small models or just RGB vs RGB+Audio in the self-supervised track?
We will ask participants to provide the total number of model parameters and the modalities used and plan to create special mentions for those doing well in each setting, but not specific tracks.