Project title

Untrimmed video human activity recognition using temporal annotation with timeline

Submitted to:

  • Disha G. Deotale
  • Dr. Madhushi Verma

Submitted By:

  1. Abbireddy sachitha
  2. Shivam  Joshi
  3. Ashish Hari S
  4. Dachepalli Mallika

Project Description

Current techniques for human action acknowledgment face numerous difficulties, for example, the requirement for various sensors, poor usage, problematic continuous execution, and absence of worldly area. In this examination, we built up a strategy for perceiving and finding human exercises dependent on worldly activity acknowledgment. Contrasted and past strategies for activity arrangement, the proposed technique includes the time limit and viably improves the identification precision. To test this technique observationally, we led tests using the games recordings, for example, cricket, high jump and long jump. Three exercises were perceived and situated in the untrimmed long video. The precision of the outcomes demonstrated the adequacy and continuous execution of the proposed strategy, exhibiting that this methodology has incredible potential for reasonable application.

Project Poster

Poster_1
Get Latest Notification about
leadingindia.ai

Please ignore if you have already signed up.

Announcements, news and innovations!

From leadingindia.ai in your inbox.

By submitting this form, you are consenting to receive marketing emails from: Bennett University. You can revoke your consent to receive emails at any time by using the SafeUnsubscribe® link, found at the bottom of every email.