COMPUTER BASED OBJECT DETECTION WITHIN A VIDEO OR IMAGE
Described herein are software and systems for analyzing videos and/or images. Software and systems described herein are configured in different embodiments to carry out different types of analyses. For example, in some embodiments, software and systems described herein are configured to locate an object of interest within a video and/or image.
1 . (canceled)
2 . A non-transitory computer readable storage medium for identifying an object of interest or a factor of interest, the medium comprising a computer program configured to cause a processor to:
(a) input a video comprising a plurality of frames into a software module;
(b) generate a feature map from a frame of the plurality of frames with the software module, wherein the feature map comprises a probability of presence of an object of interest or factor of interest at a location within the frame; and
(c) analyzing the feature map using a statistical technique to obtain one or more probability distribution functions from the probability, thereby identifying the object of interest or factor of interest within the video.
3 . The medium of claim 2 , wherein the software module comprises a deep neural network.
4 . The medium of claim 3 , wherein the feature map comprises data from a hidden layer or an output layer of the deep neural network.
5 . The medium of claim 3 , wherein the deep neural neatwork comprises at least one of VGG-19, ResNet, Inception, and MobileNet.
6 . The medium of claim 2 , wherein the factor of interest comprises at least one of a location of a pixel within the frame and an angle within the frame.
7 . The medium of claim 2 , wherein the feature map identifies a likelihood that multiple objects of interest are located within the frame.
8 . The medium of claim 2 , wherein the statistical technique comprises Monte Carlo Sampling, and wherein the Monte Carlo sampling is used to generate sample locations of the object of interest within the feature map.
9 . The medium of claim 8 , wherein the Monte Carlo sampling is used to sample the likelihood of the presence of the object of interest for at least one of the sample locations within the frame.
10 . The medium of claim 7 , wherein the statistical technique further comprises Bayesian modeling, and wherein the Bayesian modeling is used to model a change in a location of the object of interest within the frame to a different location of the object of interest within a different frame of the plurality of frames.
11 . The medium of claim 10 , wherein the Bayesian modeling represents a set of variables and their conditional dependencies.
12 . The medium of claim 10 , wherein the computer program is further configured to cause the processor to identify a position of the object of interest within the frame relative to a different object of interest within the frame.
13 . The medium of claim 2 , wherein the factor of interest comprises an angle.
14 . The medium of claim 2 , wherein the object of interest comprises a joint of a body of an individual.
15 . The medium of claim 14 , wherein the joint comprises a shoulder, elbow, hip, knee, or ankle.
16 . The medium of claim 15 , wherein the video captures the individual within the frame.
17 . The medium of claim 16 , wherein the video captures a factor of interest from the frame to a different frame within the plurality of frames.
18 . The medium of claim 17 , wherein the factor of interest comprises a movement of a joint.
19 . The medium of claim 18 , wherein the movement of the joint is measured relative to a different joint of the body of the individual and is expressed as an angle.
20 . The medium of claim 19 , wherein the angle is used by a healthcare provider to evaluate the joint of the individual.
21 . The medium of claim 2 , wherein a Gaussian distributed heatmap is multiplied to the feature map in order to incorporate an assumption that the object of interest or factor of interest does not deviate largely between adjacent frames of the plurality of frame.