CS Objects - Jerod Weinman

	A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W	X	Y	Z	AA	AB	AC	AD	AE	AF	AG	AH	AI	AJ	AK	AL	AM	AN	AO	AP	AQ	AR	AS	AT	AU	AV	AW	AX	AY	AZ	BA
1	PID	WORKSPACE	Import_Index	PARENT	CMODEL	SEQUENCE	OBJ	TRANSCRIPT	THUMBNAIL	Title	Alternative_Titles	Personal_Names~Roles	Corporate_Names~Roles	U1_Faculty_Identifiers	U2_Department_Identifiers	Abstract	Index_Date	Date_Issued	Date_Captured	Other_Date~Display_Label	Publisher	Place_Of_Publication	Public_Notes~Types	Notes~Display_Label	Dates_as_Notes~Display_Label	Citations	Table_of_Contents	LCSH_Subjects	Subjects_Names~Types	Subjects_Geographic	Subjects_Temporal	Keywords	Coordinate	Related_Items~Type^URI	Type_of_Resource~AuthorityURI	Genre~AuthorityURI	Extent	Form~AuthorityURI	MIME_Type	Digital_Origin	Classifications~Authorities	Language_Names~Codes	Local_Identifier	Handle	Physical_Location	Shelf_Locator	Access_Condition	Import_Source	Primary_Sort	Hidden Creator	Pull_Quotes	Private_Notes~Types	PID: Copy of Column A

2	# grinnell:35245		0	grinnell:faculty-scholarship	islandora:compoundCModel	0				Descriptive Attributes for Language-based Object Keypoint Detection							2023																				6 objects
3	# grinnell:35246		1	2	islandora:sp_pdf	6	weinman23descriptive.pdf			Descriptive Attributes for Language-based Object Keypoint Detection		Weinman, Jerod (Faculty/Staff) \| Belongie, Serge \| Frank, Stella		weinman	CS	Multimodal vision and language (VL) models have recently shown strong performance in phrase grounding and object detection for both zero-shot and finetuned cases. We adapt a VL model (GLIP) for keypoint detection and evaluate on NABirds keypoints. Our languagebased keypoints-as-objects detector GLIP-KP outperforms baseline topdown keypoint detection models based on heatmaps and allows for zero and few-shot evaluation. When fully trained enhancing the keypoint names with descriptive attributes gives a significant performance boost, raising AP by as much as 6.0, compared to models without attribute information. Our model exceeds heatmap-based HRNet’s AP by 4.4 overall and 8.4 on keypoints with attributes. With limited data, attributes raise zero-/one-/few-shot test AP by 1.0/3.4/1.6, respectively, on keypoints with attributes.	2023	21-Sep-2023						September 2024~Embargo Date											research paper		14 pages			born digital		English~en
4	# grinnell:35247		2	2	islandora:binaryObjectCModel	2	nabird_parts_captions.tar.bz2			Annotations of descriptive attributes for NABirds image files using COCO JSON format		Weinman, Jerod (Faculty/Staff) \| Frank, Stella					2023																		dataset
5	# grinnell:35248		3	2	islandora:binaryObjectCModel	3	glip_attr_maxdets_1_epoch_9.pth			PyTorch (v1.10) GLIP model checkpoints for Narrow Labels+Attributes Finetune case (MaxDets=1, Epoch=9)		Weinman, Jerod (Faculty/Staff)					2023																		dataset
6	# grinnell:35249		4	2	islandora:binaryObjectCModel	4	glip_attr_maxdets_100_epoch_34.pth			PyTorch (v1.10) GLIP model checkpoints for Broad Labels+Attributes Finetune case (MaxDets=100, Epoch=34)		Weinman, Jerod (Faculty/Staff)					2023																		dataset
7	# grinnell:35250		5	2	islandora:binaryObjectCModel	5	icvs23-attributes.zip			GLIPv1-L modifications used to train the atributes-enhanced model							2023																		dataset
8	# grinnell:35251		6	2	islandora:sp_pdf	1	ICVS23_SNCS_ProceedingsPaper_LTP_ST_SN_Switzerland.pdf			Licence to Publish Proceedings Papers							2023																		document
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100