Ë
    cÔ´jG  ã                   ó²   — d Z g d¢ZdZddddddd	d
dddddddœZi dd“dd“dd“dd“dd“dd“dd“d d!“d"d#“d$d%“d&d'“d(d)“d*d+“d,d-“d.d/“d0d1“d2d3“d4d5d6d7d8d9d:œ¥Zy;)<z`Reading guides for every figure and table of the report, plus a glossary (used by f2_report.py).))ÚTrialzcOne continuous recording (50 s to 10 min) in which a listener hears two talkers and attends to one.)zColumn 0 / column 1z“The two envelope streams stored per trial. Column 0 is the stream attended for most of the trial (converter convention); column 1 is the other one.)ÚEnvelopeznThe slow loudness contour of a talker's speech (roughly 1-9 Hz). The decoder tries to reconstruct it from EEG.)zBackward decoderzËA linear filter that maps the last 0-500 ms of EEG after each moment onto the speech envelope at that moment. The talker whose envelope correlates better with the reconstruction is predicted as attended.)zr(att), r(ign), dzšCorrelation of the reconstruction with the attended and ignored envelope over a window; d = r(att) - r(ign). d > 0 means the window was decoded correctly.)ÚSNR1zšmean(d) / SD(d) over 1-s windows: how strong the 1-s evidence is compared with its noise. Accuracy follows Phi(k * SNR1 * sqrt(window)) (the SNR law, H4).)ÚLOTOzfLeave-one-trial-out within a subject: train on the subject's other trials, test on the held-out trial.)ÚLOSOzXLeave-one-subject-out within a dataset: train on the other subjects of the same dataset.)ÚLODOzºLeave-one-dataset-out: train on four datasets, test on the fifth. This is the closest proxy for the blind test sets (new subjects, new recording setup). KUL-AV-GC is the main proxy fold.)zWithin-trial splitz”Random windows split into train/test with windows of the same trial on both sides. Leaky; shown only to measure how much a decoder memorises trials.)zMedian [95% CI]zªThe median over subjects, with a 95% bootstrap confidence interval (2,000 resamples of subjects). If two CIs do not overlap, the difference is very unlikely to be chance.)zp, qz´p = Wilcoxon signed-rank test across subjects (or binomial test across trials). q = p corrected for many tests (Benjamini-Hochberg FDR). Below 0.05 is conventionally 'significant'.)z	Cohen's dzeEffect size: difference of means divided by the pooled SD. About 0.2 is small, 0.5 medium, 0.8 large.)zAcc, SDT, L, Ha!  The official metrics. Acc = % of 128 Hz samples labelled correctly. SDT = switch detection time in s (Inf = missed). L = median over switches of exp(-SDT/60), averaged over subjects (1 = instant, 0 = missed). H = harmonic mean of Acc and L; the headline score (official baseline H = 0.70).)zSYNTHETIC switchzèA switch imposed on a fixed-attention trial by swapping which stream counts as attended from a random time on (same rule as the official generateSyntheticSwitches.m). It measures reaction speed on data where no real switch happened.)Ú
Integratorz»A causal post-processor that turns the noisy 1-s evidence into stable labels: argmax (none), moving average, HMM or a state ladder. More smoothing = higher Acc but later switch detection.)zNull (sign-randomised)zjThe same pipeline fed with evidence whose sign is randomised: what a decoder with zero information scores.)ÚCARzNCommon average reference: subtracting the mean of all channels at each sample.)zEA / running EAz—Euclidean alignment: whitening each recording by its own mean covariance so recordings look alike. Running EA uses only the past (usable at test time).)zCausal / zero-phasez‘Causal filters use only past samples (required by the 1-s lookahead rule); zero-phase filters also use future samples (not allowed at test time).)zMontage 57 / 58 / 62ut   Channel sets matched by name: 57 common to all five datasets, 58 = BioSemi-64 âˆ© USTC, 62 = BioSemi-64 âˆ© KUL-UHD.a.  Colours: every figure uses one fixed colour per dataset - KUL-AV-GC red, DTU blue, KUL2016 green, KUL-UHD purple, USTC orange. Grey marks a baseline or null. Numbers are summarised per subject first and then across subjects, so a subject with many trials counts once. Unless stated otherwise, accuracies are medians over subjects with 95% bootstrap CIs in brackets, and n (subjects, trials or windows) is given in every caption. Chance level is 50% for every two-way decision. Each figure's plotted values are also saved as a CSV next to the PNG in figures/.aã  Each small panel is one quality metric; its title says which, and 'log10' means the axis is logarithmic (each unit = 10x). Each column of dots is a dataset (one dot per trial); the shaded shape (violin) shows where trials cluster and the black bar is the median. Look for datasets whose cloud sits far from the others: e.g. median_rms_sd near 0 (log10 of 1) for KUL-AV-GC and KUL2016 means no DC offset, while DTU, KUL-UHD and USTC sit at 50-200x; median_abs_dc tells the same story.aë  Power spectrum of the raw EEG: x = frequency (Hz), y = power, both on log scales. The line is the median trial of a dataset and the shaded band covers the middle 50% of trials. What to see: the red line (KUL-AV-GC) collapses below 1 Hz and above ~40 Hz (it was filtered before release); the bump near 10 Hz is the alpha rhythm (strongest in KUL2016); the sharp spike at 50 Hz is mains noise (KUL-UHD, USTC); the other datasets keep large power below 1 Hz because they were never high-passed.aÉ  Each disc is a head seen from above (nose at the top, left ear on the left); dots are electrodes. The colour shows in what share of trials that electrode was flagged bad (white = never, dark red = often). Each panel has its own colour scale (see its colour bar), so compare patterns rather than colours across panels. KUL-UHD is drawn as dots on its 253-channel layout. Clusters at the edge (e.g. frontal USTC, temporal DTU) point to poor electrode contact.a  Bars give the share of trials with a large jump (> 20 robust SD) within 0.5 s of the trial start (solid) or end (faded). A value of 1.0 means every trial. KUL-UHD and USTC reach 100%: every one of their trials needs the edges trimmed or the filters warmed up before use.a°  Each dot is one trial: the loudness (SD) of the attended envelope divided by that of the ignored one, on a log2 scale. 0 = equally loud, +1 = attended twice as loud, -1 = half as loud. The black bar is the median. A cloud shifted above 0 (KUL2016) means the attended talker is systematically louder - a shortcut. Dots lined up in horizontal rows are the same stimulus pair reused across subjects (KUL2016 has only 12 distinct rows).az  How fast the speech loudness fluctuates: x = modulation frequency (Hz), y = relative power, log scales. Curves on top of each other mean the envelopes were extracted the same way. KUL-AV-GC (red) drops sharply below 1 Hz (its envelopes were high-passed); USTC (orange) has a bump at 4-5 Hz, the syllable rate of Mandarin, and a different shape - a different envelope extraction.aõ  How strongly the EEG follows the speech envelope at each delay. x = delay in ms (positive = EEG after the sound), y = correlation strength averaged over the 57 channels (RMS). Solid = attended talker, dashed = ignored talker; the band covers the middle 50% of subjects. A peak at 100-200 ms is the brain's response to the speech; the gap between solid and dashed at the peak is the attention effect every decoder relies on. It is visible in all five datasets but small (correlations around 0.01-0.02).a£  Head maps (nose up) of the envelope-EEG correlation at each dataset's peak delay. Rows: attended talker, ignored talker, and their difference. Red = positive, blue = negative correlation; within a column the three maps share one colour scale. Where the attended map is stronger than the ignored one (strong colour in the bottom row) is where attention information lives - mostly central/frontal and temporal electrodes.af  The forward model (temporal response function): how the EEG responds over time to the speech envelope. x = delay (ms), y = response strength summed over channels (global field power). Solid = attended, dashed = ignored. Peaks between roughly 100 and 250 ms are the auditory response; a higher solid curve means attended speech drives the brain more strongly.a   Decoding accuracy (y) against the length of the decision window in seconds (x, log scale), for three validation schemes: within subject (left), new subject from the same dataset (middle) and new dataset (right). Points are the median over subjects, bars the 95% CI; the grey line is chance (0.5). All curves rise with window length (more data per decision = fewer errors). The right panel is the realistic one for the challenge: about 53-56% at 1 s and 56-67% at 10 s. DTU has no 60-s point because its trials are only 48 s long after trimming.a  Test of the SNR law. Each dot is one subject at one window length; x = the accuracy predicted from that subject's 1-s SNR alone, y = the accuracy actually measured. Dots on the dashed diagonal mean the prediction is exact. The legend gives the fitted k per dataset (1 = consecutive seconds independent; below 1 = they are correlated, so longer windows help slightly less than ideal).aõ  Each dot is one 5-s EEG segment, placed by the two main directions of variation of its spatial covariance (axes have no units). Colour = dataset. Four panels: no alignment, scale-normalised, Euclidean alignment, causal running alignment; the title gives how accurately a classifier can tell the dataset. Separate coloured islands mean each dataset has its own 'fingerprint'; the more the colours mix after alignment, the more the alignment removed it. They mix only partly (84-87% still identifiable).a·  Signals averaged around the 99 real KUL-AV-GC side switches: x = time from the switch (s; 0 = switch), y = signal level, flipped so that a left-to-right switch would push the curve upward. Left: alpha lateralisation; middle: eye-movement proxy from frontal electrodes; right: the recorded EOG channels. A physiological marker of the switch would appear as a step at 0. The curves stay flat (the band is the 95% CI), so no marker was found.ao  The accuracy-versus-reaction trade-off, one panel per held-out dataset. x = Acc (% of samples correct), y = L (switch-detection score, 1 = instant). Each coloured marker is one integrator setting (marker shape = family: circle argmax, square moving average, triangles HMM, diamond ladder); grey markers are the same settings fed with random evidence (the null). The curved grey lines are iso-H lines: every point on a line has the same headline score H, labelled on the line. Better = up and to the right. The coloured cloud sits clearly above the grey null, but far below the dotted oracle line - there is room to improve.)Úa1_qc_metric_distributionsÚa2_psd_median_iqrÚa3_bad_channel_topographyÚa4_edge_transientsÚb4_env_sd_ratioÚb7_env_modspecÚc1_tracking_curvesÚc1_topographiesÚc1_forward_trf_gfpÚc4_accuracy_vs_windowÚ
c4_snr_lawÚd1_embeddingÚd4_switch_lockedz.e1_tradeoff_c19-common57-car_lodo_pool_0-500msé   a  One row per dataset. 'Distinct audio' counts different speech recordings found by comparing envelopes, not file names - a small number means the same stories are reused across trials and subjects. 'EOG channels' says whether eye-movement electrodes were stored.é   au  One row per dataset; values are medians over subjects. Compare a column down the rows: SD and |DC| show the amplitude scale and offset (large = raw, unfiltered data); RMS/SD = 1 means no offset; 50 Hz ratio > 1 means mains noise, < 1 a notch filter; sub-1 Hz ratio < 1 means the data were high-passed; the last three columns are the share of problematic trials or channels.é   zÖThe processing each dataset most likely received before release, inferred from the spectra. Read across a row to get one dataset's history; datasets with different rows need harmonised preprocessing before pooling.é   a!  Counts from the per-sample ground-truth labels. 'Within-trial speaker switches' is 0 everywhere; only KUL-AV-GC has direction switches. 'Direction samples mislabelled' is the share of samples where the single trial-level label is wrong (50% in KUL-AV-GC because the side changes half way).é   a  Two competing explanations of the KUL-AV-GC 5-min event, tested by decoding. Each cell is the accuracy expected if that explanation were true. The 'same speaker all trial' column rises with window length (84% at 60 s) while the other falls far below 50%, so the speaker did not change.é   a:  'col0 SD > col1 SD' is the share of trials where the attended envelope is louder (50% = no bias). The p-values test whether that share differs from 50%. The last column shows how well a classifier picks the attended talker from loudness statistics alone - anything clearly above 50% is a shortcut (KUL2016: 82.5%).é   zß'Label vs trial parity agreement' of 100% means the attended side can be read off whether the trial number is odd or even. The two classifier columns show how well slow drift alone predicts the label (50% = no information).é   zÿHow many different recordings of speech exist per dataset, found from the envelopes themselves. 'Audio clusters with several names' > 0 means one recording appears under different file names - so file names cannot be trusted to build story-disjoint folds.é	   zÇFor each dataset, how often a decoder that never saw the trial prefers column 1 (the stream marked 'ignored'). Well below 50% means the columns are labelled correctly; a swap would push it above 50%.é
   a%  Size and timing of neural tracking per dataset: the delay of the peak (ms), the correlation strength for attended and ignored speech, their difference with CI, and a p-value testing that attended > ignored across subjects. All p-values are small, so the attention effect is present everywhere.é   zíAccuracy (%) for the three validation schemes (rows) at 1, 10 and 60-s windows, per dataset (columns). Compare down a column to see the cost of generalising: within subject > new subject > new dataset. Brackets are 95% CIs over subjects.é   zòSNR1 per scheme (rows) and dataset (columns). Higher = stronger evidence per second. The three 'lodo' rows are different ways of pooling training data; they are almost identical. Compare with 0.099, the value implied by the official baseline.é   ai  Each row is one change to the preprocessing, compared with the first row (reference). The number is the median accuracy; in brackets the median change vs the reference in percentage points and its p-value. Negative changes mean the variant is worse. Nothing beats the causal reference with a significant margin except, marginally, the 0.5-30 Hz variant at 10 s.é   a  How well a classifier can tell which dataset (or which subject) a 5-s EEG segment came from. Rows: without and with alignment. Chance for dataset ID is 31% (always guessing the largest dataset). Values near 100% mean a strong fingerprint; alignment lowers it only to 84-87%.é   ar  Direction (left/right) decoding with spatial filters. Read along a row from left to right: the leaky within-trial split is highest, then accuracy falls as training and test data get further apart (held-out trial, new subject, new dataset). The last column shows the same features identifying the trial itself - far above chance - which is what inflates the first column.é   a  Does brain or eye activity differ between attending right and left? d is the effect size per subject (median, 95% CI); 'share positive' is the share of subjects with d > 0; q is the p-value corrected for all tests in the table. Only USTC's eye channel (eog0) shows a clear effect.é   z³The change from 60 s before to 60 s after each real KUL-AV-GC switch (median over subjects, 95% CI) and its p-value. Values near 0 with large p mean nothing changes at the switch.zÈHow the allowed 1-s lookahead is spent: the decoder looks 500 ms into the future of each sample, and decisions are used 500 ms earlier than the end of their window. The total must not exceed 1,000 ms.am  Switch-detection scores per held-out dataset for the best integrator. Acc, SDT (seconds to detect), L and H are the official metrics; 'miss rate' = share of switches never detected. The 'null' columns are the same pipeline with random evidence, i.e. the floor. 'Oracle misses' counts switches that even a perfect decoder loses because of the scorer's 50-s property.z¨The best H each integrator family reaches on each held-out dataset. Compare down a column: smoothing integrators (moving average, HMM) beat plain argmax by 0.02-0.07 H.zdOne row per hypothesis: what it claims, why it matters for the challenge and how it was tested here.z|One row per hypothesis: the claim, the verdict (colour-coded in the key above) and a plain-language summary of the evidence.z�The same verdicts with the full numbers: the previous (single-trial) estimate next to the full-corpus estimate, and the sample size n behind it.)é   é   é   é   é   é   N)Ú__doc__ÚGLOSSARYÚCONVENTIONSÚ	FIG_HOWTOÚ	TAB_HOWTO© ó    ú]/root/disk_3tb/EEG-sAAD-Grand-Challenge-Training-Dataset/reports/eda_full/scripts/f2_howto.pyÚ<module>r6      s±  ðÙ fò$€ðN7ð ð	vð
	{ð
	bð
	 ð	8ð
	Fð	Bð
	3ð
	lð	-ð	Ið	|ð
	Að
	@ñAE€	ðN8Øð Oð8ð ð Eð8ð ð gð8ð ð -ð8ð ð 3ð8ð ð Lð8ð" ð jð#8ð& ð ð'8ð, ð Sð-8ð0 ð 	6ð18ð6 ð 	xð78ð: ð 	|ð;8ð> ð 	zð?8ðD ð 	ðE8ðJ ð 	IðK8ðP ð 	/ðQ8ðV ð 	BðW8ðZ	Tð	~ð	:ànð 	Gð	òm8�	r4   