High Number of Missing Frames Observed

[D-Idan]

Issue: Missing Frames in RADDet Dataset
Hi,

I'm working with the RADDet dataset and have encountered an issue with missing frames. I'm primarily using the data from OpenDataLab (https://opendatalab.com/OpenDataLab/RADDet), as the OneDrive and Google Drive links seem to be unavailable.

Specifically, I've noticed that some sequences appear to have missing frames, which is causing problems with my analysis. I'm trying to understand if this is a known issue with the dataset, or if there might be a problem with the data on OpenDataLab.

My questions are:

Is the issue of missing frames a known problem with the RADDet dataset?

Are there any other reliable sources for the complete RADDet dataset, given that the OneDrive and Google Drive links are not currently working?

Any insights or guidance on this would be greatly appreciated.

Thank you!

[D-Idan]

Here is a summary report of the results for frame sampling at 10 Hz, presented for both the train set and test set, with different max_gap values and sorted by length:

---

📊 Summary Table

🧪 Train Set

max_gap	Longest Sequence (Frames)	Number of Files	Duration (seconds)
0	007808 → 007848	41	4.1 s
1	002830 → 002968	125	12.5 s
2	006275 → 006787	416	41.6 s

---

🧪 Test Set

max_gap | Longest Sequence (Frames) | Number of Files | Duration (seconds) -- | -- | -- | -- 0 | 004905 → 004910 | 6 | 0.6 s 1 | 004353 → 004361 | 7 | 0.7 s 2 | 006908 → 006931 | 13 | 1.3 s

[AnuvabSen1]

Did the issue get solved?