Scenic route-based navigation directions

1. A method for selecting a scenic route in response to a request for navigation directions, the method comprising:

training, by one or more processors, a machine learning model using (i) map feature data for each of a plurality of road segments, and (ii) a scenic metric for each of the plurality of road segments indicative of a quality of a view from the road segment including classifying each of the plurality of road segments into a subset of a plurality of subsets of road segments each having a corresponding scenic metric or range of scenic metrics, and analyzing the map feature data for each subset of road segments to generate the machine learning model;

receiving, by the one or more processors, a request for navigation directions from a starting location to a destination location;

identifying, by the one or more processors, a set of candidate routes for navigating from the starting location to the destination location;

for each road segment on each candidate route in the set of candidate routes:

obtaining, by the one or more processors, map feature data for the road segment; and

applying, by the one or more processors, the machine learning model to the map feature data for the road segment to generate a scenic metric for the road segment;

selecting, by the one or more processors, a route from the set of candidate routes based at least in part on the scenic metrics for each candidate route; and

providing, by the one or more processors, a set of navigation directions for presentation on a client device for navigating from the starting location to the destination location via the selected route.

2. The method of claim 1 , further comprising:

training, by the one or more processors, the machine learning model further based on a date or time in which the scenic metric is obtained for each road segment; and

applying, by the one or more processors, a date or time in which the navigation directions are requested to the machine learning model to generate the scenic metric for the road segment.

3. The method of claim 1 , further comprising:

training, by the one or more processors, the machine learning model further based on one or more terms associated with each road segment; and

for each road segment on each candidate route in the set of candidate routes:

obtaining, by the one or more processors, one or more terms associated with the road segment; and

applying, by the one or more processors, the one or more terms to the machine learning model to generate the scenic metric for the road segment.

4. The method of claim 1 , further comprising:

obtaining, by the one or more processors, for at least one of the road segments on at least one candidate route in the set of candidate routes, a visibility metric indicative of visibility of a landmark from the road segment; and

adjusting, by the one or more processors, the scenic metric based on the obtained indication.

5. The method of claim 1 , wherein selecting a route from the set of candidate routes based at least in part on the scenic metrics for each candidate route includes:

for each candidate route:

assigning, by the one or more processors, a score to each road segment within the candidate route based at least in part on the scenic metric of the road segment;

combining, by the one or more processors, the scores for each road segment within the candidate route to generate an overall score; and

selecting, by the one or more processors, the route from the set of candidate routes according to the overall score for each respective candidate route.

6. The method of claim 5 , wherein assigning a score to each road segment within the candidate route based at least in part on the scenic metric of the road segment includes:

assigning, by the one or more processors, a score to each road segment within the candidate route based on an estimated amount of time for traversing the road segment; and

adjusting, by the one or more processors, the score for each road segment based on the scenic metric for the road segment.

7. The method of claim 1 , wherein the map feature data for each of the plurality of road segments includes at least one of: a location of the road segment, a distance from the road segment to a nearest landmark, a size of the nearest landmark, a visibility metric indicative of visibility of the nearest landmark from the road segment, or a popularity metric indicative of popularity of the landmark.

8. The method of claim 1 , wherein receiving a request for navigation directions from a starting location to a destination location includes receiving a request for a tour around a geographic area beginning and ending at the starting location, and further comprising:

for each road segment in the geographic area:

obtaining, by the one or more processors, map feature data for the road segment; and

applying, by the one or more processors, the machine learning model to the map feature data for the road segment to generate a scenic metric for the road segment; and

selecting, by the one or more processors, a route based at least in part on the scenic metrics for each road segment in the geographic area.

9. The method of claim 8 , wherein selecting a route based at least in part on the scenic metrics for each road segment in the geographic area includes:

identifying, by the one or more processors, a set of road segments each having a scenic metric above a threshold metric; and

selecting, by the one or more processors, a route including each of the set of road segments and one or more additional road segments for traversing the set of road segments beginning and ending at the starting location.

10. A computing device for selecting a scenic route in response to a request for navigation directions, the computing device comprising:

one or more processors; and

a non-transitory computer-readable memory coupled to the one or more processors and storing instructions thereon that, when executed by the one or more processors, cause the computing device to:

train a machine learning model using (i) map feature data for each of a plurality of road segments, and (ii) a scenic metric for each of the plurality of road segments indicative of a quality of a view from the road segment, including classifying each of the plurality of road segments into a subset of a plurality of subsets of road segments each having a corresponding scenic metric or range of scenic metrics, and analyzing the map feature data for each subset of road segments to generate the machine learning model;

receive a request for navigation directions from a starting location to a destination location;

identify a set of candidate routes for navigating from the starting location to the destination location;

for each road segment on each candidate route in the set of candidate routes:

obtain map feature data for the road segment; and

apply the machine learning model to the map feature data for the road segment to generate a scenic metric for the road segment;

select a route from the set of candidate routes based at least in part on the scenic metrics for each candidate route; and

provide a set of navigation directions for presentation on a client device for navigating from the starting location to the destination location via the selected route.

11. The computing device of claim 10 , wherein the instructions further cause the computing device to:

train the machine learning model further based on a date or time in which the scenic metric is obtained for each road segment; and

apply a date or time in which the navigation directions are requested to the machine learning model to generate the scenic metric for the road segment.

12. The computing device of claim 10 , wherein the instructions further cause the computing device to:

train the machine learning model further based on one or more terms associated with each road segment; and

for each road segment on each candidate route in the set of candidate routes:

obtain one or more terms associated with the road segment; and

apply the one or more terms to the machine learning model to generate the scenic metric for the road segment.

13. The computing device of claim 10 , wherein the instructions further cause the computing device to:

obtain for at least one of the road segments on at least one candidate route in the set of candidate routes, a visibility metric indicative of visibility of a landmark from the road segment; and

adjust the scenic metric based on the obtained indication.

14. The computing device of claim 10 , wherein to select a route from the set of candidate routes based at least in part on the scenic metrics for each candidate route, the instructions cause the computing device to:

for each candidate route:

assign a score to each road segment within the candidate route based at least in part on the scenic metric of the road segment;

combine the scores for each road segment within the candidate route to generate an overall score; and

select the route from the set of candidate routes according to the overall score for each respective candidate route.

15. The computing device of claim 14 , wherein to assign a score to each road segment within the candidate route based at least in part on the scenic metric of the road segment, the instructions cause the computing device to:

assign a score to each road segment within the candidate route based on an estimated amount of time for traversing the road segment; and

adjust the score for each road segment based on the scenic metric for the road segment.

16. The computing device of claim 10 , wherein the map feature data for each of the plurality of road segments includes at least one of: a location of the road segment, a distance from the road segment to a nearest landmark, a size of the nearest landmark, a visibility metric indicative of visibility of the nearest landmark from the road segment, or a popularity metric indicative of popularity of the landmark.

17. The computing device of claim 10 , wherein the request for navigation directions from a starting location to a destination location is a request for a tour around a geographic area beginning and ending at the starting location, and wherein the instructions further cause the computing device to:

for each road segment in the geographic area:

obtain map feature data for the road segment; and

apply the machine learning model to the map feature data for the road segment to generate a scenic metric for the road segment; and

select a route based at least in part on the scenic metrics for each road segment in the geographic area.

18. A method for providing navigation directions in response to a request for a tour of a geographic area, the method comprising:

receiving, by the one or more processors, a request for a tour of a geographic area beginning and ending at a starting location;

for each road segment in the geographic area, assigning, by the one or more processors, a scenic metric to the road segment indicative of a quality of a view from the road segment by applying characteristics of the road segment to a machine learning model trained to generate scenic metrics for road segments;

selecting, by the one or more processors, a subset of the road segments in the geographic area based on the scenic metrics of the road segments;

generating, by the one or more processors, a route beginning and ending at the starting location, the route including each of the selected subset of road segments; and

providing, by the one or more processors, a set of navigation directions for presentation on a client device for navigating on the tour around the geographic area via the generated route.

19. The method of claim 18 , wherein selecting the subset of the road segments includes selecting, by the one or more processors, the subset of road segments in the geographic area each having a scenic metric above a threshold metric.

20. The method of claim 18 , wherein selecting the route includes selecting, by the one or more processors, the route including each of the selected subset of road segments and one or more additional road segments for traversing the set of road segments beginning and ending at the starting location.