Coral Cover: Chagos #2 2015

Shaded value = in the CV mean (shade deepens with cover) Gray italic = shown for comparison but left out of the CV mean: the gold standard (the reference value) and the manually annotated in CoralNet CoralNet benchmark annotations (reference values, not models), Coralscapes (not trusted), and any value that reuses the reference's own data n = frames behind the value — = not run or not applicable (see each region's notes) Site names and method names link out (new tab)

Coral Cover: Chagos #2 2015

49 sites · 1,132 photos

Gold standard = LOF diver transects (no photos). 480 photos manually scored (via CoralNet).

Site	Date	Headline (site page)	Claude	Gold standard	Manually annotated in CoralNet	CoralNet	DINOv3	Coralscapes v1	Coralscapes v2	CoralSCOP	Mean ± SD of CV	Range
BIPB60	2015-04-14	42%transect	30^a,in=38	42^a,b	39^a,cn=38	—	33^a,fn=38	55^a,gn=38	57^a,gn=38	42^a,hn=38	35 ± 6.3	30–42
BIPB61	2015-04-14	20%transect	17^a,in=14	20^a,b	18^a,cn=14	—	22^a,fn=14	41^a,gn=14	43^a,gn=14	32^a,hn=14	24 ± 7.9	17–32
BIPB62	2015-04-14	59%transect	34^a,in=19	59^a,b	48^a,cn=19	—	35^a,fn=19	58^a,gn=18	58^a,gn=18	32^a,hn=19	33 ± 1.8	32–35
BIPB63	2015-04-15	56%transect	35^a,in=29	56^a,b	47^a,cn=29	—	33^a,fn=29	66^a,gn=29	67^a,gn=29	38^a,hn=29	36 ± 2.6	33–38
BIPB64	2015-04-15	33%transect	23^a,in=21	33^a,b	23^a,cn=21	—	26^a,fn=21	34^a,gn=21	41^a,gn=21	33^a,hn=21	27 ± 5.4	23–33
BIPB65	2015-04-15	47%transect	30^a,in=19	47^a,b	33^a,cn=19	—	32^a,fn=19	52^a,gn=19	55^a,gn=19	28^a,hn=19	30 ± 2.4	28–32
BIPB66	2015-04-16	7.8%transect	too few photos (n=6)^a,i,d	7.8^a,b	too few photos (n=6)^a,c,d	—	too few photos (n=6)^a,f,d	too few photos (n=6)^a,g,d	too few photos (n=6)^a,g,d	too few photos (n=6)^a,h,d	—	—
BIPB67	2015-04-16	46%transect	26^a,in=41	46^a,b	36^a,cn=41	—	34^a,fn=41	49^a,gn=41	52^a,gn=41	29^a,hn=41	30 ± 3.9	26–34
BIPB68	2015-04-16	59%transect	37^a,in=43	59^a,b	45^a,cn=43	—	39^a,fn=43	72^a,gn=43	74^a,gn=43	39^a,hn=43	38 ± 1.2	37–39
BIPB69	2015-04-16–05-02	73%transect	39^a,in=116	73^a,b	54^a,cn=110	too few photos (n=6)^a,e,d	47^a,fn=116	74^a,gn=116	75^a,gn=116	43^a,hn=116	43 ± 4.2	39–47
BIPB70	2015-04-17–05-01	70%transect	35^a,in=55	70^a,b	45^a,cn=28	45^a,en=27	40^a,fn=55	57^a,gn=55	59^a,gn=55	36^a,hn=55	39 ± 4.7	35–45
BIPB72	2015-04-18–05-01	77%transect	44^a,in=75	77^a,b	51^a,cn=41	56^a,en=34	49^a,fn=75	64^a,gn=75	69^a,gn=75	40^a,hn=75	47 ± 6.9	40–56
BIPB73	2015-04-18	50%transect	38^a,in=45	50^a,b	—	45^a,en=45	45^a,fn=45	48^a,gn=45	51^a,gn=45	47^a,hn=45	44 ± 4.2	38–47
BIPB74	2015-04-18–05-01	63%transect	51^a,in=21	63^a,b	too few photos (n=8)^a,c,d	56^a,en=13	49^a,fn=21	64^a,gn=21	66^a,gn=21	47^a,hn=21	51 ± 3.8	47–56
BIPB75	2015-04-19	46%transect	too few photos (n=9)^a,i,d	46^a,b	—	too few photos (n=9)^a,e,d	too few photos (n=9)^a,f,d	too few photos (n=9)^a,g,d	too few photos (n=9)^a,g,d	too few photos (n=9)^a,h,d	—	—
BIPB76	2015-04-19	40%transect	21^a,in=20	40^a,b	—	30^a,en=20	32^a,fn=20	35^a,gn=20	40^a,gn=20	29^a,hn=20	28 ± 4.7	21–32
BINI77	2015-04-20	42%transect	26^a,in=11	42^a,b	—	30^a,en=11	26^a,fn=11	38^a,gn=11	41^a,gn=11	34^a,hn=11	29 ± 3.7	26–34
BINI78	2015-04-20	22%transect	25^a,in=16	22^a,b	—	24^a,en=16	29^a,fn=16	27^a,gn=16	29^a,gn=16	28^a,hn=16	27 ± 2.4	24–29
BINI79	2015-04-20	34%transect	18^a,in=15	34^a,b	—	19^a,en=15	22^a,fn=15	27^a,gn=15	25^a,gn=15	21^a,hn=15	20 ± 1.7	18–22
BINI80	2015-04-21	49%transect	29^a,in=24	49^a,b	38^a,cn=10	27^a,en=14	37^a,fn=24	42^a,gn=24	46^a,gn=24	35^a,hn=24	32 ± 4.7	27–37
BIVB81	2015-04-21	43%transect	too few photos (n=5)^a,i,d	43^a,b	—	too few photos (n=5)^a,e,d	too few photos (n=5)^a,f,d	too few photos (n=5)^a,g,d	too few photos (n=5)^a,g,d	too few photos (n=5)^a,h,d	—	—
BIBL82	2015-04-22	42%transect	28^a,in=22	42^a,b	too few photos (n=4)^a,c,d	28^a,en=18	26^a,fn=22	42^a,gn=22	50^a,gn=22	36^a,hn=22	29 ± 4.3	26–36
BIBL83	2015-04-22	33%transect	24^a,in=15	33^a,b	too few photos (n=9)^a,c,d	too few photos (n=6)^a,e,d	25^a,fn=15	42^a,gn=15	45^a,gn=15	41^a,hn=15	30 ± 9.5	24–41
BISA84	2015-04-22	49%transect	too few photos (n=7)^a,i,d	49^a,b	too few photos (n=2)^a,c,d	too few photos (n=5)^a,e,d	too few photos (n=7)^a,f,d	too few photos (n=7)^a,g,d	too few photos (n=7)^a,g,d	too few photos (n=7)^a,h,d	—	—
BIBL86	2015-04-23	43%transect	34^a,in=27	43^a,b	—	32^a,en=27	36^a,fn=27	51^a,gn=27	49^a,gn=27	25^a,hn=27	32 ± 4.8	25–36
BIBL87	2015-04-23	43%transect	30^a,in=35	43^a,b	—	28^a,en=35	34^a,fn=35	42^a,gn=35	48^a,gn=35	29^a,hn=35	30 ± 2.6	28–34
BISA88	2015-04-23	60%transect	43^a,in=13	60^a,b	—	46^a,en=13	54^a,fn=13	55^a,gn=13	60^a,gn=13	43^a,hn=13	46 ± 5.2	43–54
BISA91	2015-04-24	57%transect	40^a,in=13	57^a,b	—	47^a,en=13	56^a,fn=13	58^a,gn=13	65^a,gn=13	44^a,hn=13	47 ± 6.5	40–56
BISP89	2015-04-24	34%transect	20^a,in=13	34^a,b	—	17^a,en=13	22^a,fn=13	29^a,gn=13	30^a,gn=13	28^a,hn=13	22 ± 4.4	17–28
BISP90	2015-04-24	44%transect	too few photos (n=8)^a,i,d	44^a,b	—	too few photos (n=8)^a,e,d	too few photos (n=8)^a,f,d	too few photos (n=8)^a,g,d	too few photos (n=8)^a,g,d	too few photos (n=8)^a,h,d	—	—
BISA92	2015-04-25	26%transect	19^a,in=39	26^a,b	—	29^a,en=39	26^a,fn=39	33^a,gn=39	34^a,gn=39	29^a,hn=39	26 ± 4.9	19–29
BISA93	2015-04-25	41%transect	33^a,in=27	41^a,b	—	39^a,en=27	26^a,fn=28	58^a,gn=26	61^a,gn=27	26^a,hn=28	31 ± 5.9	26–39
BISA94	2015-04-25	52%transect	too few photos (n=2)^a,i,d	52^a,b	—	too few photos (n=2)^a,e,d	too few photos (n=2)^a,f,d	too few photos (n=2)^a,g,d	too few photos (n=2)^a,g,d	too few photos (n=2)^a,h,d	—	—
BISA97	2015-04-26	55%transect	51^a,in=12	55^a,b	—	51^a,en=12	50^a,fn=12	68^a,gn=12	69^a,gn=12	29^a,hn=12	45 ± 10.6	29–51
BISP95	2015-04-26	61%transect	18^a,in=26	61^a,b	—	22^a,en=26	20^a,fn=26	28^a,gn=26	28^a,gn=26	23^a,hn=25	21 ± 2.3	18–23
BISP96	2015-04-26	82%transect	51^a,in=18	82^a,b	—	56^a,en=18	49^a,fn=18	69^a,gn=18	73^a,gn=18	48^a,hn=18	51 ± 3.3	48–56
BIVB99	2015-04-27	60%transect	35^a,in=11	60^a,b	—	21^a,en=11	25^a,fn=11	64^a,gn=11	63^a,gn=11	19^a,hn=11	25 ± 7.5	19–35
BISA102	2015-04-28	32%transect	22^a,in=27	32^a,b	—	31^a,en=27	30^a,fn=27	33^a,gn=27	35^a,gn=27	34^a,hn=27	29 ± 5.5	22–34
BISP100	2015-04-28	41%transect	too few photos (n=7)^a,i,d	41^a,b	—	too few photos (n=7)^a,e,d	too few photos (n=7)^a,f,d	too few photos (n=7)^a,g,d	too few photos (n=7)^a,g,d	too few photos (n=7)^a,h,d	—	—
BISP101	2015-04-28	52%transect	23^a,in=13	52^a,b	—	39^a,en=13	32^a,fn=13	42^a,gn=13	44^a,gn=13	33^a,hn=13	32 ± 6.6	23–39
BISA103	2015-04-29	33%transect	30^a,in=19	33^a,b	—	28^a,en=19	37^a,fn=19	41^a,gn=19	44^a,gn=19	36^a,hn=19	33 ± 4.4	28–37
BISA104	2015-04-29	36%transect	36^a,in=16	36^a,b	—	44^a,en=16	46^a,fn=16	46^a,gn=16	51^a,gn=16	41^a,hn=16	42 ± 4.4	36–46
BISA105	2015-04-29	52%transect	28^a,in=19	52^a,b	—	24^a,en=19	36^a,fn=19	42^a,gn=19	47^a,gn=19	23^a,hn=19	28 ± 5.8	23–36
BIPB110	2015-04-30	42%transect	29^a,in=45	42^a,b	—	37^a,en=45	37^a,fn=45	44^a,gn=43	50^a,gn=43	37^a,hn=44	35 ± 3.6	29–37
BISA107	2015-04-30	25%transect	18^a,in=12	25^a,b	—	18^a,en=12	21^a,fn=12	32^a,gn=12	41^a,gn=12	22^a,hn=12	20 ± 1.8	18–22
BIVB108	2015-04-30	47%transect	12^a,in=11	47^a,b	too few photos (n=1)^a,c,d	25^a,en=10	24^a,fn=11	23^a,gn=10	18^a,gn=11	28^a,hn=11	22 ± 7.2	12–28
BIVB109	2015-04-30	48%transect	35^a,in=33	48^a,b	too few photos (n=1)^a,c,d	35^a,en=32	39^a,fn=33	67^a,gn=33	66^a,gn=33	25^a,hn=33	33 ± 6.0	25–39
BIPB112	2015-05-02	46%transect	29^a,in=14	46^a,b	44^a,cn=14	—	44^a,fn=14	51^a,gn=14	55^a,gn=14	38^a,hn=14	37 ± 7.8	29–44
BIPB114	2015-05-02–03	43%transect	20^a,in=29	43^a,b	29^a,cn=22	too few photos (n=7)^a,e,d	32^a,fn=29	37^a,gn=29	38^a,gn=29	35^a,hn=29	29 ± 8.0	20–35
All sites (unweighted mean)		—	—	46	—	—	—	—	—	—	—	—

Notes for Chagos #2 2015 (13)

[a] Value from the map's computer-vision method tables (denominator and photo set as on the site's page).
[b] Gold = LOF in-water diver transect, the mean of all of the station's transects: a different instrument, site-level only, never within 200 m of the frames.
[c] The author's own CoralNet points on the ground-truth frames at this site (30 − WATER − FISH, unknown points kept).
[d] Too few photos (under 10) at this site for a value: not shown and not averaged (photo-count rule: 10 or more photos for a value, 20 or more for a confident one).
[e] CoralNet robot labels (the cover set), same denominator.
[f] DINOv3 point predictions at the same CoralNet points (2,500 frames).
[g] Coralscapes (v1 EPFL-ECEO/coralscapes-vit-b-dpt; v2 coralscapesv2-dinov3-vitb-lora-dpt) is not considered reliable for these photos: shown, never averaged. Area fraction, not a point count.
[h] CoralSCOP = per-frame sum of colony-mask area fractions (zero-filled; overlapping masks double-count), averaged over frames with an ok row. Area index, not a point count: denominator differs from any point-count reference. It is also unstable frame to frame (the same subject has scored 94.1% in one photo and 23.3% in another), and the mask sum overstates cover most on the healthiest reefs.
[i] Claude run complete (2,500 of 2,500 frames; 2,495 assigned to these 104 site rows): blind calls at 30 CoralNet points on the EDITED website photos, HC+BL / (30 − WATER − FISH), unknown points kept.
Mean/Range: Mean ± SD and Range are over the CV cells that enter the mean (CoralNet where independent, DINOv3, CoralSCOP unedited/edited, Claude). Gold standard, frames manually annotated in CoralNet, Coralscapes and same-measurement cells are shown in gray italic but never averaged.
Headline (site page): the coral-cover figure the site's page, report card and map pop-up print, with its method; it is never averaged into the CV mean.
Methods: Claude (Anthropic); DINOv3 (Meta; paper; used here as the frozen backbone of the in-house classifier); Coralscapes v1 and v2 (EPFL-ECEO; dataset paper, code); CoralSCOP (Zheng et al., CVPR 2024); CoralNet.
Sites: names drop the region prefix (hover for the full name) and link to the site's CRD page, or to the regional summary page where the site has none. Date = survey date from the map pins (first–last visit); sites are listed earliest first, undated sites last.

All regions