Datasets:

leideng
/

cockpit-vit-images

image_id stringlengths 10 13	source stringclasses 1 value	height int32 108 1.5k	width int32 116 1.5k	image_bytes int64 2.89k 545k
ai2d/row-307	ai2d	120	350	6,388
ai2d/row-2624	ai2d	350	750	25,723
ai2d/row-271	ai2d	403	548	40,581
ai2d/row-736	ai2d	390	409	27,674
ai2d/row-1081	ai2d	318	400	25,643
ai2d/row-466	ai2d	327	529	28,766
ai2d/row-2544	ai2d	882	1,024	94,486
ai2d/row-2742	ai2d	370	876	41,323
ai2d/row-1343	ai2d	297	464	25,106
ai2d/row-2211	ai2d	442	681	47,742
ai2d/row-2788	ai2d	450	560	35,272
ai2d/row-1258	ai2d	450	680	51,093
ai2d/row-1168	ai2d	193	219	8,137
ai2d/row-2803	ai2d	600	800	56,198
ai2d/row-1564	ai2d	844	1,500	183,609
ai2d/row-2718	ai2d	546	728	85,414
ai2d/row-1833	ai2d	343	298	17,738
ai2d/row-919	ai2d	154	268	6,659
ai2d/row-955	ai2d	224	332	15,865
ai2d/row-485	ai2d	699	768	87,313
ai2d/row-393	ai2d	587	934	60,003
ai2d/row-3006	ai2d	353	322	15,528
ai2d/row-492	ai2d	540	600	39,151
ai2d/row-2308	ai2d	320	504	14,610
ai2d/row-1186	ai2d	586	882	60,933
ai2d/row-820	ai2d	746	1,277	79,125
ai2d/row-41	ai2d	390	246	18,667
ai2d/row-2881	ai2d	184	474	19,553
ai2d/row-1422	ai2d	394	600	46,896
ai2d/row-1284	ai2d	510	910	61,239
ai2d/row-1450	ai2d	546	728	41,172
ai2d/row-2336	ai2d	1,463	1,438	175,119
ai2d/row-731	ai2d	223	284	14,558
ai2d/row-2789	ai2d	314	466	20,156
ai2d/row-598	ai2d	497	312	25,198
ai2d/row-2407	ai2d	466	550	45,041
ai2d/row-2708	ai2d	308	521	10,044
ai2d/row-208	ai2d	479	638	69,299
ai2d/row-516	ai2d	846	947	101,329
ai2d/row-1328	ai2d	197	300	11,967
ai2d/row-2008	ai2d	257	328	32,127
ai2d/row-1431	ai2d	180	219	3,408
ai2d/row-180	ai2d	380	576	58,731
ai2d/row-1513	ai2d	599	800	70,923
ai2d/row-256	ai2d	799	1,213	137,904
ai2d/row-850	ai2d	507	564	40,520
ai2d/row-1650	ai2d	448	434	39,637
ai2d/row-3013	ai2d	954	599	97,914
ai2d/row-2855	ai2d	314	466	20,156
ai2d/row-481	ai2d	591	837	75,565
ai2d/row-2886	ai2d	231	350	11,967
ai2d/row-2858	ai2d	662	612	42,112
ai2d/row-2594	ai2d	375	600	36,182
ai2d/row-452	ai2d	1,500	1,500	305,510
ai2d/row-1098	ai2d	688	591	43,670
ai2d/row-832	ai2d	382	500	53,108
ai2d/row-126	ai2d	599	484	29,489
ai2d/row-2669	ai2d	558	460	15,683
ai2d/row-300	ai2d	546	598	45,548
ai2d/row-2425	ai2d	449	338	14,920
ai2d/row-2082	ai2d	699	768	87,313
ai2d/row-3060	ai2d	612	792	52,237
ai2d/row-1610	ai2d	403	562	44,496
ai2d/row-11	ai2d	539	508	45,524
ai2d/row-286	ai2d	300	300	19,272
ai2d/row-2737	ai2d	1,500	625	108,942
ai2d/row-69	ai2d	471	500	54,543
ai2d/row-1148	ai2d	224	332	15,865
ai2d/row-656	ai2d	200	420	18,054
ai2d/row-744	ai2d	760	550	48,324
ai2d/row-1815	ai2d	401	398	22,649
ai2d/row-1033	ai2d	429	479	43,184
ai2d/row-703	ai2d	293	400	38,000
ai2d/row-2220	ai2d	720	960	89,288
ai2d/row-105	ai2d	322	449	26,229
ai2d/row-501	ai2d	353	236	19,099
ai2d/row-392	ai2d	268	306	9,922
ai2d/row-842	ai2d	266	359	27,112
ai2d/row-2490	ai2d	172	300	12,554
ai2d/row-1955	ai2d	231	476	13,516
ai2d/row-319	ai2d	480	480	40,687
ai2d/row-1546	ai2d	507	564	40,520
ai2d/row-2647	ai2d	250	300	13,848
ai2d/row-1277	ai2d	258	456	21,310
ai2d/row-2880	ai2d	720	960	73,352
ai2d/row-662	ai2d	768	1,024	73,295
ai2d/row-2965	ai2d	1,500	1,500	396,348
ai2d/row-2446	ai2d	800	836	48,216
ai2d/row-704	ai2d	375	500	23,819
ai2d/row-1486	ai2d	546	728	85,414
ai2d/row-815	ai2d	412	600	45,811
ai2d/row-2589	ai2d	600	633	47,943
ai2d/row-1320	ai2d	724	800	83,462
ai2d/row-966	ai2d	353	322	15,528
ai2d/row-3008	ai2d	503	1,500	49,730
ai2d/row-535	ai2d	1,082	1,500	126,925
ai2d/row-862	ai2d	363	862	43,054
ai2d/row-937	ai2d	428	236	26,691
ai2d/row-503	ai2d	219	500	24,853
ai2d/row-168	ai2d	408	550	24,096

End of preview. Expand in Data Studio

Cockpit ViT Images

Reproducible image pool for ViT token compression. Each config is one upstream corpus. Splits are train / validation / test. Rows store the image as parquet-embedded bytes plus image_id, source, subset, height, width, and image_bytes (encoded size).

Upstream licenses stay with the original datasets (COCO, Open Images, Visual Genome, MMMU, ChartQA, TextVQA, AI2D). Redistribute and cite accordingly.

Downloads last month: 1,030