Filter
Exclude
Time range
-
Near
💎 You earn points everywhere. But do you know where they actually are? Real-time tracking separates control from chaos. How important is seeing your rewards instantly?💐 #LYYL #RealTimeTracking #PointTracking #Web3Loyalty How important is real-time point tracking?
0% A) Crucial 🧨
0% B) Nice to have 🥳
0% C) Not important 😞
0% D) Neutral 👌
0 votes • 5 days
26
Only 2 weeks left to submit to the VOTS2026 challenges! Details: votchallenge.net/vots2026/pa… #VOTS #Tracking #PointTracking #VideoSegmentation

🎉 The VOTS2026 challenge is now open! 🏁Submit to 4 tracks: VOTS2026, VOTSp2026, VOTSr2026, VOTSt2026! 🏆Challenges close on June 22nd. ℹ️More info: votchallenge.net/vots2026/ #VOT #VideoSegmentation #ECCV2026
38
10 Oct 2025
Let your camera follow your art. Literally. With edelkrone’s exclusive Point Tracking, you can set multiple tracking points and switch between them effortlessly—unlocking fluid, cinematic motion that keeps up with your creative flow. Perfect for solo artists who need more than static shots. #edelkrone #motioncontrol #sliderplus #solocreator #creativeprocess #pointtracking #filmmakinggear #contentcreation #headplus #slidemasterbundle
1
4
429
Cubsat "HORIZON" in PointTracking mode for our Mission Control Center made several images and sent them to Earth. A rope, which used to hold the antennas, but is now in the drift.
23
691
🚀 ProTracker delivers robust Tracking Any Point (TAP) with a Kalman filter-inspired Probabilistic approach, seamlessly fusing optical flow and semantic cues for smoother, more accurate trajectories! ✨It sets the SOTA among self-supervised methods—outperforming even supervised ones on several benchmarks—and excels in occlusion, low-feature areas, and challenging scenarios! 💡🔗 Check it out! #PointTracking #ComputerVision #CV #DeepLearning #AI #Trackinganypoint #TAP #Reconstruction #Vision #ML #MachineLearning Project page: michaelszj.github.io/protrac… Paper: arxiv.org/abs/2501.03220 Code: michaelszj.github.io/protrac…

🚀Introducing Protracker: Inspired by Kalman filter, we tackle point tracking with a robust probabilistic approach. 🌟Our method integrates multiple predictions from both optical flow and semantic correspondences in a unified framework with probabilistic fusion. This ensures to generate smooth and accurate trajectories. ProTracker achieves state-of-the-art performance among self-supervised methods across multiple benchmarks and enhanced robustness in challenging scenarios like occlusion, similar regions, and low-feature areas. 💡Protracker offers a probabilistic framework that combines information of different granularity and semantics, paving the way for new advancements in tracking any point. 🔍#PointTracking #ComputerVision Project page: michaelszj.github.io/protrac… Paper: arxiv.org/abs/2501.03220 Code: michaelszj.github.io/protrac…
2
19
2,371
🚀Introducing Protracker: Inspired by Kalman filter, we tackle point tracking with a robust probabilistic approach. 🌟Our method integrates multiple predictions from both optical flow and semantic correspondences in a unified framework with probabilistic fusion. This ensures to generate smooth and accurate trajectories. ProTracker achieves state-of-the-art performance among self-supervised methods across multiple benchmarks and enhanced robustness in challenging scenarios like occlusion, similar regions, and low-feature areas. 💡Protracker offers a probabilistic framework that combines information of different granularity and semantics, paving the way for new advancements in tracking any point. 🔍#PointTracking #ComputerVision Project page: michaelszj.github.io/protrac… Paper: arxiv.org/abs/2501.03220 Code: michaelszj.github.io/protrac…
3
11
85
11,850
COTRACKER3: SIMPLER AND BETTER POINT TRACKING BY PSEUDO-LABELLING REAL VIDEOS: Meta's CoTracker3 simplifies the process of tracking points across videos by using a streamlined architecture and an efficient pseudo-labelling approach with minimal data. Key Highlights: ✅ Data Efficiency: With only 15k real videos, CoTracker3 achieves state-of-the-art results, even outperforming models like BootsTAPIR, which used 15 million videos! This model is proof that efficiency in design can deliver exceptional performance. ✅ Handling Occlusions with Ease: Using cross-track attention, CoTracker3 accurately predicts the locations of occluded points based on visible ones, making it especially powerful in complex tracking scenarios. ✅ Self-Training Advantage: CoTracker3 fine-tunes on its own predictions, gaining an impressive performance boost ( 1.2 points on TAP-Vid benchmarks) and reducing the gap between synthetic and real-world data. ✅ Top Speed and Efficiency: CoTracker3 processes frames 30% faster than LocoTrack, the previous speed leader, while maintaining high accuracy—even with large numbers of tracked points. ✅ Scalability with Minimal Data: Its performance scales effectively with just a moderate increase in training data (up to 30k videos), showing excellent balance in training requirements without compromising results. ✅ The authors also discuss some failure cases, providing insights into areas where CoTracker3 could improve further. This model has exciting applications in areas like 3D tracking, real-time video analysis, and dynamic 3D reconstruction. Project Page: cotracker3.github.io/ Paper: arxiv.org/pdf/2410.11831 Github: github.com/facebookresearch/… @jianyuan_wang @n_karaev #ComputerVision #Tracking #Research #PointTracking #CoTracker #LongTracking #Meta #DenseTracking
2
4
227
With ~500 loyalty programs, 75 trackable, Oro's got you covered! #oro #digitalcards #smartwallet #mtl #pointtracking
1