egrace479 commited on
Commit
91e4716
·
verified ·
1 Parent(s): ec7d4c4

Add some more clarification around links and direct to kabr-tools docs

Browse files
Files changed (1) hide show
  1. README.md +6 -7
README.md CHANGED
@@ -17,13 +17,13 @@ tags:
17
  - UAV
18
  - drone
19
  - video
20
- model_description: "Behavior recognition model for in situ drone videos of zebras and giraffes, built using X3D model initialized on Kinetics weights. It is trained on the KABR dataset, which is comprised of 10 hours of aerial video footage of reticulated giraffes (Giraffa reticulata), Plains zebras (Equus quagga), and Grevy’s zebras (Equus grevyi) captured using a DJI Mavic 2S drone. It includes both spatiotemporal (i.e., mini-scenes) and behavior annotations provided by an expert behavioral ecologist."
21
  ---
22
 
23
  # Model Card for X3D-KABR-Kinetics
24
  X3D-KABR-Kinetics is a behavior recognition model for in situ drone videos of zebras and giraffes,
25
  built using X3D model initialized on Kinetics weights.
26
- It is trained on the [KABR dataset](https://huggingface.co/datasets/imageomics/KABR),
27
  which is comprised of 10 hours of aerial video footage of reticulated giraffes
28
  (*Giraffa reticulata*), Plains zebras (*Equus quagga*), and Grevy’s zebras
29
  (*Equus grevyi*) captured using a DJI Mavic 2S drone.
@@ -50,7 +50,7 @@ This model was developed for the benefit of the community as an open-source prod
50
 
51
  - **Repository:** [Project Repo](https://github.com/Imageomics/kabr-tools)
52
  - **Paper:** [Paper Link](https://link.springer.com/article/10.1007/s11042-024-20512-4)
53
- - **Project Page** [Project Page](https://imageomics.github.io/KABR/)
54
 
55
  ## Uses
56
 
@@ -58,7 +58,7 @@ X3D-KABR-Kinetics has extensively studied ungulate behavior classification from
58
 
59
  ### Direct Use
60
 
61
- Please see the illustrative examples on the [kabr-tools repository](https://github.com/Imageomics/kabr-tools)
62
  for more information on how this model can be used generate time-budgets from aerial video of animals.
63
 
64
  ### Out-of-Scope Use
@@ -81,8 +81,7 @@ This model was trained to detect and classify behavior from drone videos of zebr
81
 
82
  ## How to Get Started with the Model
83
 
84
-
85
- Please see the illustrative examples on the [kabr-tools repository](https://github.com/Imageomics/kabr-tools)
86
  for more information on how this model can be used generate time-budgets from aerial video of animals.
87
 
88
  ## Training Details
@@ -124,7 +123,7 @@ The dataset was evaluated on the X3D-L model utilizing the [SlowFast](https://gi
124
 
125
  ### Testing Data
126
 
127
- We provide a train-test split of the mini-scenes from the [KABR Dataset](https://huggingface.co/datasets/imageomics/KABR) for evaluation purposes (test set indicated in [annotations/val.csv](https://huggingface.co/datasets/imageomics/KABR/blob/main/KABR/annotation/val.csv), with 75% for train and 25% for testing. No mini-scene was divided by the split. The splits ensured a stratified representation of giraffes, Plains zebras, and Grevy’s zebras.
128
 
129
  #### Metrics
130
 
 
17
  - UAV
18
  - drone
19
  - video
20
+ model_description: "Behavior recognition model for in situ drone videos of zebras and giraffes, built using X3D model initialized on Kinetics weights. It is trained on the KABR mini-scene dataset, which is comprised of 10 hours of aerial video footage of reticulated giraffes (Giraffa reticulata), Plains zebras (Equus quagga), and Grevy’s zebras (Equus grevyi) captured using a DJI Mavic 2S drone. It includes both spatiotemporal (i.e., mini-scenes) and behavior annotations provided by an expert behavioral ecologist."
21
  ---
22
 
23
  # Model Card for X3D-KABR-Kinetics
24
  X3D-KABR-Kinetics is a behavior recognition model for in situ drone videos of zebras and giraffes,
25
  built using X3D model initialized on Kinetics weights.
26
+ It is trained on the [KABR mini-scene dataset](https://huggingface.co/datasets/imageomics/KABR),
27
  which is comprised of 10 hours of aerial video footage of reticulated giraffes
28
  (*Giraffa reticulata*), Plains zebras (*Equus quagga*), and Grevy’s zebras
29
  (*Equus grevyi*) captured using a DJI Mavic 2S drone.
 
50
 
51
  - **Repository:** [Project Repo](https://github.com/Imageomics/kabr-tools)
52
  - **Paper:** [Paper Link](https://link.springer.com/article/10.1007/s11042-024-20512-4)
53
+ - **Project Page:** [KABR Project Page](https://imageomics.github.io/KABR/), [KABR mini-scene project page](https://kabrdata.xyz/)
54
 
55
  ## Uses
56
 
 
58
 
59
  ### Direct Use
60
 
61
+ Please see the illustrative examples in the [kabr-tools docs](https://imageomics.github.io/kabr-tools/case_studies/overview/)
62
  for more information on how this model can be used generate time-budgets from aerial video of animals.
63
 
64
  ### Out-of-Scope Use
 
81
 
82
  ## How to Get Started with the Model
83
 
84
+ Please see the illustrative examples in the [kabr-tools docs](https://imageomics.github.io/kabr-tools/)
 
85
  for more information on how this model can be used generate time-budgets from aerial video of animals.
86
 
87
  ## Training Details
 
123
 
124
  ### Testing Data
125
 
126
+ We provide a train-test split of the mini-scenes from the [KABR mini-scene dataset](https://huggingface.co/datasets/imageomics/KABR) for evaluation purposes (test set indicated in [annotations/val.csv](https://huggingface.co/datasets/imageomics/KABR/blob/main/KABR/annotation/val.csv), with 75% for train and 25% for testing. No mini-scene was divided by the split. The splits ensured a stratified representation of giraffes, Plains zebras, and Grevy’s zebras.
127
 
128
  #### Metrics
129