Commits · Luigi/Real-Time-Faint-Detection-on-Video

increase image editor zone

abb9c26

Luigi commited on Apr 16

adjust example path

5b4536d

Luigi commited on Apr 16

add a royality-free example video

aca8c5c

Luigi commited on Apr 16

use mid-hip to determine if a person is inside of alert zone but keep using bottom centor for velocity estimation

f881a88

Luigi commited on Apr 16

show skeleton

fc238fe

Luigi commited on Apr 16

use mid-hip to detct if a person appear inside of alert zone instead of using bottom center which may lead to some false-alarm

bdb253c

Luigi commited on Apr 16

Apply low-pass filtering on velocity estimator output

0adc388

Luigi commited on Apr 16

adjust motionless timer lower limit in ui

8566a51

Luigi commited on Apr 15

adjust ui laytout

95a4d18

Luigi commited on Apr 15

Use Yolov11s pose to unify all models

efdfcf3

Luigi commited on Apr 15

enabel zerogpu

f8e2b6e

Luigi commited on Apr 15

show alert zone in red line

be9331b

Luigi commited on Apr 15

add alert zone define zone, and detc faint only in the zone

65abf59

Luigi commited on Apr 15

remove environmental classification

6163bb1

Luigi commited on Apr 15

decrease gpu dur to 200 sec

baa9eed

Luigi commited on Apr 15

increase zero gpu duration

38ccaf3

Luigi commited on Apr 15

improve annotation layout

744f3b5

Luigi commited on Apr 15

Analysiss result readibility improvement

a2ce116

Luigi commited on Apr 15

bugfix

c5c4fd1

Luigi commited on Apr 14

velocity threshold +environmental classification result

cc42f27

Luigi commited on Apr 14

add 2nd method based on velocity for motionlessness detection

5357618

Luigi commited on Apr 14

apply rtdetr_r50vd_coco_o365 for exnvironmental classification

a2dbbf9

Luigi commited on Apr 14

fix model id for rt-detr

39b2b6d

Luigi commited on Apr 14

add context awareness by environmental classification

b11c742

Luigi commited on Apr 14

parse vitpose output

bf6d73d

Luigi commited on Apr 14

do kpt post processing diy

5eaad40

Luigi commited on Apr 14

bugfix to ViTPose processing error: Batch of images and boxes mismatch

d50f53a

Luigi commited on Apr 14

add missing arg to image_processor call

831fc31

Luigi commited on Apr 14

add debug info

a865cc7

Luigi commited on Apr 14

rescaled pose measue to be aligned with bbox scale

c974a8e

Luigi commited on Apr 14

again, tuen para in is_lying_from_pose

e949ed0

Luigi commited on Apr 14

tune parameter in is_lying_from_pose

bd8314d

Luigi commited on Apr 14

undebug

df3bb3f

Luigi commited on Apr 14

debug with vitpose bypassed

aee935c

Luigi commited on Apr 14

back to vitpose small

e9011e0

Luigi commited on Apr 14

yolov11s

2bf2707

Luigi commited on Apr 14

pure yolov5

23ed1d9

Luigi commited on Apr 14

switch to yolov5

058a48c

Luigi commited on Apr 14

fix for zero gpu

4518ae4

Luigi commited on Apr 14

fix for vitpose plus

ee4ea03

Luigi commited on Apr 14

vitpose-small

ec7b69b

Luigi commited on Apr 14

bacl to yolov8n for pd

48d17a2

Luigi commited on Apr 14

use vitpose for od & kd

4b6d068

Luigi commited on Apr 14

switch to medium model for od

0242402

Luigi commited on Apr 14

add zero gpu support

8b12e6b

Luigi commited on Apr 14

initial commit

19d7478

Luigi commited on Apr 14

Commit History

increase image editor zone abb9c26

adjust example path 5b4536d

add a royality-free example video aca8c5c

use mid-hip to determine if a person is inside of alert zone but keep using bottom centor for velocity estimation f881a88

show skeleton fc238fe

use mid-hip to detct if a person appear inside of alert zone instead of using bottom center which may lead to some false-alarm bdb253c

Apply low-pass filtering on velocity estimator output 0adc388

adjust motionless timer lower limit in ui 8566a51

adjust ui laytout 95a4d18

Use Yolov11s pose to unify all models efdfcf3

enabel zerogpu f8e2b6e

show alert zone in red line be9331b

add alert zone define zone, and detc faint only in the zone 65abf59

remove environmental classification 6163bb1

decrease gpu dur to 200 sec baa9eed

increase zero gpu duration 38ccaf3

improve annotation layout 744f3b5

Analysiss result readibility improvement a2ce116

bugfix c5c4fd1

velocity threshold +environmental classification result cc42f27

add 2nd method based on velocity for motionlessness detection 5357618

apply rtdetr_r50vd_coco_o365 for exnvironmental classification a2dbbf9

fix model id for rt-detr 39b2b6d

add context awareness by environmental classification b11c742

parse vitpose output bf6d73d

do kpt post processing diy 5eaad40

bugfix to ViTPose processing error: Batch of images and boxes mismatch d50f53a

add missing arg to image_processor call 831fc31

add debug info a865cc7

rescaled pose measue to be aligned with bbox scale c974a8e

again, tuen para in is_lying_from_pose e949ed0

tune parameter in is_lying_from_pose bd8314d

undebug df3bb3f

debug with vitpose bypassed aee935c

back to vitpose small e9011e0

yolov11s 2bf2707

pure yolov5 23ed1d9

switch to yolov5 058a48c

fix for zero gpu 4518ae4

fix for vitpose plus ee4ea03

vitpose-small ec7b69b

bacl to yolov8n for pd 48d17a2

use vitpose for od & kd 4b6d068

switch to medium model for od 0242402

add zero gpu support 8b12e6b

initial commit 19d7478

increase image editor zone

abb9c26

adjust example path

5b4536d

add a royality-free example video

aca8c5c

use mid-hip to determine if a person is inside of alert zone but keep using bottom centor for velocity estimation

f881a88

show skeleton

fc238fe

use mid-hip to detct if a person appear inside of alert zone instead of using bottom center which may lead to some false-alarm

bdb253c

Apply low-pass filtering on velocity estimator output

0adc388

adjust motionless timer lower limit in ui

8566a51

adjust ui laytout

95a4d18

Use Yolov11s pose to unify all models

efdfcf3

enabel zerogpu

f8e2b6e

show alert zone in red line

be9331b

add alert zone define zone, and detc faint only in the zone

65abf59

remove environmental classification

6163bb1

decrease gpu dur to 200 sec

baa9eed

increase zero gpu duration

38ccaf3

improve annotation layout

744f3b5

Analysiss result readibility improvement

a2ce116

bugfix

c5c4fd1

velocity threshold +environmental classification result

cc42f27

add 2nd method based on velocity for motionlessness detection

5357618

apply rtdetr_r50vd_coco_o365 for exnvironmental classification

a2dbbf9

fix model id for rt-detr

39b2b6d

add context awareness by environmental classification

b11c742

parse vitpose output

bf6d73d

do kpt post processing diy

5eaad40

bugfix to ViTPose processing error: Batch of images and boxes mismatch

d50f53a

add missing arg to image_processor call

831fc31

add debug info

a865cc7

rescaled pose measue to be aligned with bbox scale

c974a8e

again, tuen para in is_lying_from_pose

e949ed0

tune parameter in is_lying_from_pose

bd8314d

undebug

df3bb3f

debug with vitpose bypassed

aee935c

back to vitpose small

e9011e0

yolov11s

2bf2707

pure yolov5

23ed1d9

switch to yolov5

058a48c

fix for zero gpu

4518ae4

fix for vitpose plus

ee4ea03

vitpose-small

ec7b69b

bacl to yolov8n for pd

48d17a2

use vitpose for od & kd

4b6d068

switch to medium model for od

0242402

add zero gpu support

8b12e6b

initial commit

19d7478