AIRLINK 187.09 Increased By ▲ 2.40 (1.3%)
BOP 12.69 Decreased By ▼ -0.09 (-0.7%)
CNERGY 7.58 Decreased By ▼ -0.23 (-2.94%)
FCCL 40.42 Decreased By ▼ -0.42 (-1.03%)
FFL 14.86 Decreased By ▼ -0.32 (-2.11%)
FLYNG 27.36 Increased By ▲ 0.42 (1.56%)
HUBC 131.21 Increased By ▲ 0.14 (0.11%)
HUMNL 13.26 Decreased By ▼ -0.56 (-4.05%)
KEL 4.45 Decreased By ▼ -0.06 (-1.33%)
KOSM 6.01 Decreased By ▼ -0.13 (-2.12%)
MLCF 53.16 Increased By ▲ 1.93 (3.77%)
OGDC 212.59 Increased By ▲ 0.48 (0.23%)
PACE 6.06 Decreased By ▼ -0.23 (-3.66%)
PAEL 41.94 Decreased By ▼ -0.61 (-1.43%)
PIAHCLA 15.93 Decreased By ▼ -0.58 (-3.51%)
PIBTL 9.60 Increased By ▲ 0.66 (7.38%)
POWER 11.16 Increased By ▲ 0.06 (0.54%)
PPL 173.29 Decreased By ▼ -1.71 (-0.98%)
PRL 34.13 Decreased By ▼ -0.60 (-1.73%)
PTC 23.47 Decreased By ▼ -0.47 (-1.96%)
SEARL 88.09 Decreased By ▼ -6.33 (-6.7%)
SILK 1.11 Decreased By ▼ -0.03 (-2.63%)
SSGC 32.61 Decreased By ▼ -0.50 (-1.51%)
SYM 15.53 Decreased By ▼ -1.58 (-9.23%)
TELE 7.99 Decreased By ▼ -0.26 (-3.15%)
TPLP 11.00 Decreased By ▼ -0.45 (-3.93%)
TRG 59.79 Decreased By ▼ -0.46 (-0.76%)
WAVESAPP 11.28 Decreased By ▼ -0.11 (-0.97%)
WTL 1.41 Decreased By ▼ -0.04 (-2.76%)
YOUW 3.81 Decreased By ▼ -0.12 (-3.05%)
AIRLINK 187.09 Increased By ▲ 2.40 (1.3%)
BOP 12.69 Decreased By ▼ -0.09 (-0.7%)
CNERGY 7.58 Decreased By ▼ -0.23 (-2.94%)
FCCL 40.42 Decreased By ▼ -0.42 (-1.03%)
FFL 14.86 Decreased By ▼ -0.32 (-2.11%)
FLYNG 27.36 Increased By ▲ 0.42 (1.56%)
HUBC 131.21 Increased By ▲ 0.14 (0.11%)
HUMNL 13.26 Decreased By ▼ -0.56 (-4.05%)
KEL 4.45 Decreased By ▼ -0.06 (-1.33%)
KOSM 6.01 Decreased By ▼ -0.13 (-2.12%)
MLCF 53.16 Increased By ▲ 1.93 (3.77%)
OGDC 212.59 Increased By ▲ 0.48 (0.23%)
PACE 6.06 Decreased By ▼ -0.23 (-3.66%)
PAEL 41.94 Decreased By ▼ -0.61 (-1.43%)
PIAHCLA 15.93 Decreased By ▼ -0.58 (-3.51%)
PIBTL 9.60 Increased By ▲ 0.66 (7.38%)
POWER 11.16 Increased By ▲ 0.06 (0.54%)
PPL 173.29 Decreased By ▼ -1.71 (-0.98%)
PRL 34.13 Decreased By ▼ -0.60 (-1.73%)
PTC 23.47 Decreased By ▼ -0.47 (-1.96%)
SEARL 88.09 Decreased By ▼ -6.33 (-6.7%)
SILK 1.11 Decreased By ▼ -0.03 (-2.63%)
SSGC 32.61 Decreased By ▼ -0.50 (-1.51%)
SYM 15.53 Decreased By ▼ -1.58 (-9.23%)
TELE 7.99 Decreased By ▼ -0.26 (-3.15%)
TPLP 11.00 Decreased By ▼ -0.45 (-3.93%)
TRG 59.79 Decreased By ▼ -0.46 (-0.76%)
WAVESAPP 11.28 Decreased By ▼ -0.11 (-0.97%)
WTL 1.41 Decreased By ▼ -0.04 (-2.76%)
YOUW 3.81 Decreased By ▼ -0.12 (-3.05%)
BR100 11,869 Decreased By -51.1 (-0.43%)
BR30 35,588 Decreased By -219.5 (-0.61%)
KSE100 113,252 Decreased By -532.6 (-0.47%)
KSE30 35,194 Decreased By -193.2 (-0.55%)
Technology

New AI can search people based on height, gender, clothing in surveillance videos

Take full advantage of Artificial Intelligence (AI), researchers are using it in various creative ways. This time,
Published October 24, 2018

Take full advantage of Artificial Intelligence (AI), researchers are using it in various creative ways. This time, a team has used AI to distinguish people in surveillance videos by their height, gender, and clothing color.

A team of AI researchers from India have developed a tool similar to a search engine able to find people in a video. The tool is able to search for people in surveillance footage by height, clothing color and gender.

The tool was created with the help of deep learning to train a convolutional neural network (CNN) how to recognize certain human features, known as soft biometrics, via computer vision, wrote The Next Web.

Artificial Intelligence trained to spot obesity from space

In simpler words, a person gives a video to the AI and can tell the AI specific details about the person they are looking for and it’ll search it for them. For instance, upon requesting for ‘females wearing red shirts that are 153cm tall’, the AI will narrow down people meeting the criteria from the video.

Publishing their research in arXiv, researchers claimed that the algorithm ‘correctly recovers 28 persons out of 41 in a very challenging dataset with soft biometric attributes’. The team believes that with further development, the system can become more robust and an accurate search tool.

Researchers believe that this early work could possibly show a lot of potential like helping people look for specific individuals in certain surveillance footages, such as looking for a missing individual or tracking suspected criminals.

Copyright Business Recorder, 2018

Comments

Comments are closed.