|
10 | 10 |
|
11 | 11 |  |
12 | 12 |
|
13 | | -*Example from an NSD eye-tracking run: the background is the displayed NSD imagery target, blue traces show valid gaze samples from eight 3-second viewing windows (24 seconds total), and red circles show low-velocity centers under one explicit filtering rule. The model should use the time series, not the overlay.* |
| 13 | +*Example from an NSD eye-tracking run: the background is the actual displayed NSD imagery target, and gaze is mapped into the 8.4 x 8.4 degree stimulus frame. The design file lists eight 3-second presentations of this target image; six have usable processed gaze samples in this extract. Red circles show low-velocity centers under one explicit filtering rule. The model should use the time series, not the overlay.* |
14 | 14 |
|
15 | 15 | ## Tutorial framing |
16 | 16 |
|
@@ -83,6 +83,12 @@ subset is: |
83 | 83 | - One or a few small target images from: |
84 | 84 | `s3://natural-scenes-dataset/nsddata/experiments/nsdimagery/rawtargetimages/` |
85 | 85 |
|
| 86 | +Here "repeated target image" means that the same stimulus appears multiple times |
| 87 | +within the run. In the example figure, `shared0385_nsd28752.png` is scheduled at |
| 88 | +eight separate onsets in run 2. Each onset starts a 3-second image-presentation |
| 89 | +period, followed by a 1-second rest/fixation period. These are repeated |
| 90 | +presentations of the same image, not eight different screen regions. |
| 91 | + |
86 | 92 | Do **not** download the full 37 GB `nsd_stimuli.hdf5`, all subjects, all EDF |
87 | 93 | files, or any fMRI beta files for this project. |
88 | 94 |
|
@@ -143,6 +149,8 @@ Week 1 questions: |
143 | 149 | - What is the sampling rate after preprocessing? |
144 | 150 | - Which columns represent time, x gaze, y gaze, and pupil area? |
145 | 151 | - Which file tells us when the target image is on screen? |
| 152 | +- What is a target-image presentation, and how is it different from a screen |
| 153 | + region or image file? |
146 | 154 | - What does the device-native EDF file preserve, what does the NSD `.mat` |
147 | 155 | preprocessing make easier, and what are the consequences of relying on |
148 | 156 | proprietary binary formats rather than open, documented, analysis-ready |
|
0 commit comments