Visualizing machine learning predictions of human interaction with vehicles
A computing device accesses video data displaying one or more traffic entities and generates a plurality of sequences from the video data. For each sequence, the computing device identifies a plurality of stimuli in the sequence and applies a machine learning model to generate an output describing the traffic entity. The computing device generates a data structure for storing, for each sequence, information describing the sequence and linking frame indexes of stimuli from the sequence to outputs of the machine learning model. The computing device stores the data structure in association with the video data. Responsive to receiving a selection of a sequence, the computing device loads video data for the sequence. Responsive to receiving a selection of a traffic entity within the video data, the computing device generates a graphical display element including the machine learning model output for the selected traffic entity.
1. A method comprising:
accessing video data, the video data displaying one or more traffic entities;
generating, from the video data, a plurality of sequences of video frames, each sequence of video frames corresponding to a portion of the video data displaying a traffic entity of the one or more traffic entities;
for a sequence of video frames of the plurality of sequences of video frames:
identifying a plurality of stimuli in the sequence of video frames, each stimulus including a subset of video frames of the sequence of video frames; and
applying a machine learning model to each stimulus of the plurality of stimuli, the machine learning model generating an output describing a state of mind of a traffic entity displayed in video frames of the subset of video frames;
generating a data structure storing:
information describing the sequence of video frames, and information linking frame indexes of the plurality of stimuli in the sequence of video frames to corresponding outputs of the machine learning model; and
storing the data structure in association with the video data; and
presenting, via a user interface, an output of the machine learning model for a traffic entity of interest displayed in the video data based on data structures associated with the traffic entity of interest.
2. The method of claim 1 , wherein the output of the machine learning model represents a likelihood that the traffic entity of interest will cross a street intersecting with a path of a vehicle having a position and motion corresponding to the video data.
3. The method of claim 1 , wherein the output of the machine learning model includes a likelihood that the traffic entity of interest is aware of a vehicle having a position and motion corresponding to the video data.
4. The method of claim 1 , wherein the output of the machine learning model includes a likelihood that the traffic entity of interest intends to cross a street.
5. The method of claim 1 , wherein the output of the machine learning model is a probabilistic distribution.
6. The method of claim 5 , wherein the probabilistic distribution is represented in one or more graphs, charts, and tables summarizing the outputs of the machine learning model for video frames in the sequence of video frames.
7. A non-transitory computer readable medium storing instructions that when executed by one or more computer processors cause the one or more computer processors to perform steps comprising:
accessing video data, the video data displaying one or more traffic entities;
generating, from the video data, a plurality of sequences of video frames, each sequence of video frames corresponding to a portion of the video data displaying a traffic entity of the one or more traffic entities;
for a sequence of video frames of the plurality of sequences of video frames:
identifying a plurality of stimuli in the sequence of video frames, each stimulus including a subset of video frames of the sequence of video frames; and
applying a machine learning model to each stimulus of the plurality of stimuli, the machine learning model generating an output describing a state of mind of a traffic entity displayed in video frames of the subset of video frames;
generating a data structure storing:
information describing the sequence of video frames, and
information linking frame indexes of the plurality of stimuli in the sequence of video frames to corresponding outputs of the machine learning model; and
storing the data structure in association with the video data; and
presenting, via a user interface, an output of the machine learning model for a traffic entity of interest displayed in the video data based on data structures associated with the traffic entity of interest.
8. The non-transitory computer readable medium of claim 7 , wherein the output of the machine learning model represents a likelihood that the traffic entity of interest will cross a street intersecting with a path of a vehicle having a position and motion corresponding to the video data.
9. The non-transitory computer readable medium of claim 7 , wherein the output of the machine learning model includes a likelihood that the traffic entity of interest is aware of a vehicle having a position and motion corresponding to the video data.
10. The non-transitory computer readable medium of claim 7 , wherein the output of the machine learning model includes a likelihood that the traffic entity of interest intends to cross a street.
11. The non-transitory computer readable medium of claim 7 , wherein the output of the machine learning model is a probabilistic distribution.
12. The non-transitory computer readable medium of claim 11 , wherein the probabilistic distribution is represented in one or more graphs, charts, and tables summarizing the outputs of the machine learning model for video frames in the sequence of video frames.
13. A computer system comprising:
one or more computer processors; and
a non-transitory computer readable medium storing instructions that when executed by the one or more computer processors cause the one or more computer processors to perform steps comprising:
accessing video data, the video data displaying one or more traffic entities;
generating, from the video data, a plurality of sequences of video frames, each sequence of video frames corresponding to a portion of the video data displaying a traffic entity of the one or more traffic entities;
for a sequence of video frames of the plurality of sequences of video frames:
identifying a plurality of stimuli in the sequence of video frames, each stimulus including a subset of video frames of the sequence of video frames; and
applying a machine learning model to each stimulus of the plurality of stimuli, the machine learning model generating an output describing a state of mind of a traffic entity displayed in video frames of the subset of video frames;
generating a data structure storing:
information describing the sequence of video frames, and
information linking frame indexes of the plurality of stimuli in the sequence of video frames to corresponding outputs of the machine learning model; and
storing the data structure in association with the video data; and
presenting, via a user interface, an output of the machine learning model for a traffic entity of interest displayed in the video data based on data structures associated with the traffic entity of interest.
14. The computer system of claim 13 , wherein the output of the machine learning model represents a likelihood that the traffic entity of interest will cross a street intersecting with a path of a vehicle having a position and motion corresponding to the video data.
15. The computer system of claim 13 , wherein the output of the machine learning model includes a likelihood that the traffic entity of interest is aware of a vehicle having a position and motion corresponding to the video data.
16. The computer system of claim 13 , wherein the output of the machine learning model includes a likelihood that the traffic entity of interest intends to cross a street.
17. The computer system of claim 13 , wherein the output of the machine learning model is a probabilistic distribution.
18. The computer system of claim 17 , wherein the probabilistic distribution is represented in one or more graphs, charts, and tables summarizing the outputs of the machine learning model for video frames in the sequence of video frames.