ICME 2024 GC PAIR Competition Final Rankings

Dear participants,

 

Greetings, and thank you for joining us in the Low-power Efficient and Accurate Facial-Landmark Detection for Embedded Systems Competition. Over the course of this competition, we’ve seen an impressive showcase of skills and innovation from each of the 165 teams that joined us. Your efforts in navigating the complexities of facial-landmark detection on embedded systems have been truly commendable.

Facial-landmark detection, a crucial technology in computer vision, finds its importance magnified in applications ranging from augmented reality to emotion analysis. The challenge of ensuring precision and efficiency, especially on embedded systems with limited computational power, has been at the forefront of this competition.

Following an intense qualification round and a rigorous final competition evaluated on the MediaTek Dimensity Series 9300 platform, we’re moving towards the climax of this event. The competition comprised two stages: the initial model submissions for online evaluation and the final showdown where the top 15 teams, chosen based on their accuracy, demonstrated their models’ capabilities in real-world scenarios.

It’s important to note that the final accuracy submissions were held to the standard that they could not be 5% lower than the initial qualified submissions. This was to ensure consistency and reliability in performance under the varying conditions of the competition.

Please be reminded that the scores from the finals, while crucial, are not the sole determinants of the final rankings. The comprehensive evaluation will include factors such as paper review outcomes and the completeness of documentation submitted. Our aim is to recognize not just the technical excellence but also the holistic contribution to the field.

Award recipients, please note: your presence at the IEEE ICME2024 Grand Challenge PAIR Competition Special Session is essential. This session is not just a requirement but an opportunity to showcase your work and insights to a broader audience. Failure to attend will necessitate the cancellation of the award.

The final award winners will be announced shortly, and we are excited to celebrate the achievements that have emerged from this challenging yet inspiring journey.

Thank you once again for your dedication and participation. Your contributions are pushing the boundaries of what’s possible in facial-landmark detection and embedded AI technologies.

 

Best regards,

 

ICME2024 GC PAIR Competition Organizing Committee