All Projects → SUTDCV → UAV-Human

SUTDCV / UAV-Human

Licence: other
[CVPR2021] UAV-Human: A Large Benchmark for Human Behavior Understanding with Unmanned Aerial Vehicles

Programming Languages

python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to UAV-Human

MSAF
Offical implementation of paper "MSAF: Multimodal Split Attention Fusion"
Stars: ✭ 47 (-61.48%)
Mutual labels:  action-recognition
DroNet
DroNet: Efficient convolutional neural network detector for Real-Time UAV applications
Stars: ✭ 54 (-55.74%)
Mutual labels:  uav
public regulated data types
Regulated DSDL definitions for Cyphal (standard and third-party)
Stars: ✭ 62 (-49.18%)
Mutual labels:  uav
ZeroPilot-SW
Software for WARG custom autopilot
Stars: ✭ 13 (-89.34%)
Mutual labels:  uav
ViCC
[WACV'22] Code repository for the paper "Self-supervised Video Representation Learning with Cross-Stream Prototypical Contrasting", https://arxiv.org/abs/2106.10137.
Stars: ✭ 33 (-72.95%)
Mutual labels:  action-recognition
ailia-models
The collection of pre-trained, state-of-the-art AI models for ailia SDK
Stars: ✭ 1,102 (+803.28%)
Mutual labels:  action-recognition
panther
Perception-Aware Trajectory Planner in Dynamic Environments
Stars: ✭ 115 (-5.74%)
Mutual labels:  uav
gzsl-od
Out-of-Distribution Detection for Generalized Zero-Shot Action Recognition
Stars: ✭ 47 (-61.48%)
Mutual labels:  action-recognition
Action-Localization
Action-Localization, Atomic Visual Actions (AVA) Dataset
Stars: ✭ 22 (-81.97%)
Mutual labels:  action-recognition
dynamic-images-for-action-recognition
A public Python implementation for generating Dynamic Images introduced in 'Dynamic Image Networks for Action Recognition' by Bilen et al.
Stars: ✭ 27 (-77.87%)
Mutual labels:  action-recognition
mrasl mav traj
Trajectory utilities for MAVs
Stars: ✭ 26 (-78.69%)
Mutual labels:  uav
FMT-Firmware
FMT Autopilot Embedded System
Stars: ✭ 207 (+69.67%)
Mutual labels:  uav
MTL-AQA
What and How Well You Performed? A Multitask Learning Approach to Action Quality Assessment [CVPR 2019]
Stars: ✭ 38 (-68.85%)
Mutual labels:  action-recognition
FUEL
An Efficient Framework for Fast UAV Exploration
Stars: ✭ 450 (+268.85%)
Mutual labels:  uav
tfvaegan
[ECCV 2020] Official Pytorch implementation for "Latent Embedding Feedback and Discriminative Features for Zero-Shot Classification". SOTA results for ZSL and GZSL
Stars: ✭ 107 (-12.3%)
Mutual labels:  action-recognition
MUSES
[CVPR 2021] Multi-shot Temporal Event Localization: a Benchmark
Stars: ✭ 51 (-58.2%)
Mutual labels:  action-recognition
Openpose-based-GUI-for-Realtime-Pose-Estimate-and-Action-Recognition
GUI based on the python api of openpose in windows using cuda10 and cudnn7. Support body , hand, face keypoints estimation and data saving. Realtime gesture recognition is realized through two-layer neural network based on the skeleton collected from the gui.
Stars: ✭ 69 (-43.44%)
Mutual labels:  action-recognition
fixed-wing-sim
Matlab implementation to simulate the non-linear dynamics of a fixed-wing unmanned areal glider. Includes tools to calculate aerodynamic coefficients using a vortex lattice method implementation, and to extract longitudinal and lateral linear systems around the trimmed gliding state.
Stars: ✭ 72 (-40.98%)
Mutual labels:  uav
TCE
This repository contains the code implementation used in the paper Temporally Coherent Embeddings for Self-Supervised Video Representation Learning (TCE).
Stars: ✭ 51 (-58.2%)
Mutual labels:  action-recognition
Squeeze-and-Recursion-Temporal-Gates
Code for : [Pattern Recognit. Lett. 2021] "Learn to cycle: Time-consistent feature discovery for action recognition" and [IJCNN 2021] "Multi-Temporal Convolutions for Human Action Recognition in Videos".
Stars: ✭ 62 (-49.18%)
Mutual labels:  action-recognition

UAV-Human

Official repository for CVPR2021: UAV-Human: A Large Benchmark for Human Behavior Understanding with Unmanned Aerial Vehicles

Paper

[CVF OpenAccess] [arXiv] [ResearchGate]

Dataset Download

The dataset is available for Download now!

FAQs:

Q1: Is my competition result in MMVRAC comparable with the results reported in your original paper?

A1: No. During our ICCVW2021 MMVRAC competition, only part of the testing data is released, and therefore the result obtained in the MMVRAC competition is NOT comparable with the results reported in our paper. If you try to publish an academic paper on our benchmark, please follow the above link and download the whole released dataset.

Q2: Which identity should I use to train my ReID model, the person ID or the setup ID?

A2: We have provided person IDs and setup IDs.

In our original paper, we concatenated the person IDs and subject IDs as the identities to train our model. However, we provide person ID and setup ID to give more choices for you to train and evaluate your own models.

Also note that we obtained consent forms from all captured subjects.

Q3: In your ReID dataset, is it the case that you combined the subject ids with the setup ids to obtain 1,144 identities?

A3: Yes. In order to publish our ReID dataset to the public, we have to obtain the consent forms from all the captured subjects, and then we are allowed to distribute the videos containing the captured subjects to the community.

Action Classes

The UAV-Human dataset contains 155 action classes, and the action categories are listed below:

  • A000: drink
  • A001: eat snacks
  • A002: brush hair
  • A003: drop something
  • A004: pick up something
  • A005: throw away something
  • A006: sit down
  • A007: stand up
  • A008: applaud
  • A009: read
  • A010: write
  • A011: put on a coat
  • A012: take off a coat
  • A013: put on glasses
  • A014: take off glasses
  • A015: put on a hat
  • A016: take off a hat
  • A017: throw away a hat
  • A018: cheer
  • A019: wave hands
  • A020: kick something
  • A021: reach into pockets
  • A022: jump on single leg
  • A023: jump on two legs
  • A024: make a phone call
  • A025: play with cell phones
  • A026: point somewhere
  • A027: look at the watch
  • A028: rub hands
  • A029: bow
  • A030: shake head
  • A031: salute
  • A032: cross palms together
  • A033: cross arms in front to say no
  • A034: wear headphones
  • A035: take off headphones
  • A036: make a shh sign
  • A037: touch the hair
  • A038: thumb up
  • A039: thumb down
  • A040: make an OK sign
  • A041: make a victory sign
  • A042: punch with fists
  • A043: figure snap
  • A044: open the bottle
  • A045: smell
  • A046: squat
  • A047: apply cream to face
  • A048: apply cream to hands
  • A049: grasp a bag
  • A050: put down a bag
  • A051: put something into a bag
  • A052: take something out of a bag
  • A053: open a box
  • A054: move a box
  • A055: put up hands
  • A056: put hands on hips
  • A057: wrap arms around
  • A058: shake arms
  • A059: step on the spot walk
  • A060: kick aside
  • A061: kick backward
  • A062: cough
  • A063: sneeze
  • A064: yawn
  • A065: blow nose
  • A066: stagger
  • A067: headache
  • A068: chest discomfort
  • A069: backache
  • A070: neck-ache
  • A071: vomit
  • A072: use a fan
  • A073: stretch body
  • A074: punching someone
  • A075: kicking someone
  • A076: pushing someone
  • A077: slap someone on the back
  • A078: point someone
  • A079: hug
  • A080: give something to someone
  • A081: steal something from other’s pocket
  • A082: rob something from someone
  • A083: shake hands
  • A084: walk toward someone
  • A085: walk away from someone
  • A086: hit someone with something
  • A087: threat some with a knife
  • A088: bump into someone
  • A089: walk side by side
  • A090: high five
  • A091: drink a toast
  • A092: move something with someone
  • A093: take a phone for someone
  • A094: stalk someone
  • A095: whisper in someone’s ear
  • A096: exchange something with someone
  • A097: lend an arm to support someone
  • A098: rock-paper-scissors
  • A099: hover
  • A100: land
  • A101: land at designated locations
  • A102: move forward
  • A103: move backward
  • A104: move left
  • A105: move right
  • A106: ascend
  • A107: descend
  • A108: accelerate
  • A109: decelerate
  • A110: come over here
  • A111: stay where you are
  • A112: rear right turn
  • A113: rear left turn
  • A114: abandon landing
  • A115: all clear
  • A116: not clear
  • A117: have command
  • A118: follow me
  • A119: turn left
  • A120: turn right
  • A121: throw litter
  • A122: dig a hole
  • A123: mow
  • A124: set on fire
  • A125: smoke
  • A126: cut the tree
  • A127: fishing
  • A128: pick a lock
  • A129: pollute walls
  • A130: hold someone hostage
  • A131: threat someone with a gun
  • A132: wave a goodbye
  • A133: chase someone
  • A134: comfort someone
  • A135: drag someone
  • A136: sweep the floor
  • A137: mop the floor
  • A138: bounce the ball
  • A139: shoot at the basket
  • A140: swing the racket
  • A141: leg pressing
  • A142: escape (to survive)
  • A143: call for help
  • A144: wear a mask
  • A145: take off a mask
  • A146: bend arms around someone’s shoulder
  • A147: run
  • A148: stab someone with a knife
  • A149: throw a frisbee
  • A150: carry a carrying pole
  • A151: use a lever to lift something
  • A152: walk
  • A153: open an umbrella
  • A154: close an umbrella

Annotations

VideoNames: P070S01G10B00H10UC102000LC092000A031R0_09131758.avi

P070: (PersonID) unique person ID for the main subject in current video

S01: (SetupID) setup id that indicates changes of clothing, hat, backpack of the main subject

G10: (Gender) first bit represents main subject's gender, second bit represents auxiliary subject's gender

0: n/a; 1: male; 2: female

B00: (Backpack) first bit represents main subject's backpack color, second bit represents auxiliary subject's backpack color

0: n/a; 1: red; 2: black; 3: green; 4: yellow; 5: n/a

H10: (Hat) first bit represents main subject's hat color, second bit represents auxiliary subject's hat color

0: n/a; 1: red; 2: black; 3: yellow; 4: white; 5: n/a

UC102000: (UpperClothing) first three bits represent main subject's upper clothing color (2 bits) and style (1 bit), last three bits represent auxiliary subject's upper clothing color (2 bits) and style (1 bit)

color: 0: n/a; 1: red; 2: black; 3: blue; 4: green; 5: multicolor; 6: grey; 7: white; 8: yellow; 9: dark brown; 10: purple; 11: pink

style: 0: n/a; 1: long; 2: short; 3: skirt

LC102000: (LowerClothing) first three bits are main subject's lower clothing color (2 bits) and style (1 bit), last three bits are auxiliary subject's lower clothing color (2 bits) and style (1 bit)

color: 0: n/a; 1: red; 2: black; 3: blue; 4: green; 5: multicolor; 6: grey; 7: white; 8: yellow; 9: dark brown; 10: purple; 11: pink

style: 0: n/a; 1: long; 2: short; 3: skirt

A031: (Action) action labels of current sample

R00: (Replicate) replicate capturing

09131758: capturing timestamp, month(2 bits)/day(2 bits)/hour(2 bits)/minute(2 bits)

Action Recognition Evaluation Protocols

Cross-Subject-v1

In cross-subject-v1 evaluation, we split 119 subjects into training and testing groups. The IDs of training subjects are 0, 2, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 25, 26, 27, 28, 29, 30, 32, 33, 34, 35, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 55, 56, 57, 59, 61, 62, 63, 64, 65, 67, 68, 69, 70, 71, 73, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 98, 100, 102, 103, 105, 106, 110, 111, 112, 114, 115, 116, 117, 118; the remaining subjects are for testing.

Cross-Subject-v2

In cross-subject-v2 evaluation, we split 119 subjects into training and testing groups. The IDs of training subjects are 0, 3, 4, 5, 6, 8, 10, 11, 12, 14, 16, 18, 19, 20, 21, 22, 24, 26, 29, 30, 31, 32, 35, 36, 37, 38, 39, 40, 43, 44, 45, 46, 47, 49, 52, 54, 56, 57, 59, 60, 61, 62, 63, 64, 66, 67, 69, 70, 71, 72, 73, 74, 75, 77, 78, 79, 80, 81, 83, 84, 86, 87, 88, 89, 91, 92, 93, 94, 95, 96, 97, 99, 100, 101, 102, 103, 104, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 117, 118; the remaining subjects are for testing.

Results

Evaluation on I3D

Modality CSv1 - Acc (%) CSv2 - Acc (%)
RGB Video 23.86 29.53
Fisheye Video 20.76 34.12

Evaluation on skeleton

Methods CSv1 - Acc (%) CSv2 - Acc (%)
DGNN 29.90 -
ST-GCN 30.25 56.14
2s-AGCN 34.84 66.68
HARD-Net 36.97 -
Shift-GCN 37.98 67.04

Video Processing Guidance

Please refer to utils/convert_videos_to_frames.py and following:

python convert_videos_to_frames.py --videos path/to/all/videos --frames path/to/output/frames

Citation

@InProceedings{Li_2021_CVPR,
    author    = {Li, Tianjiao and Liu, Jun and Zhang, Wei and Ni, Yun and Wang, Wenqian and Li, Zhiheng},
    title     = {{UAV-Human: A Large Benchmark for Human Behavior Understanding With Unmanned Aerial Vehicles}},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2021},
    pages     = {16266-16275}
}

Contact

tianjiao_li [at] mymail.sutd.edu.sg

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].