I'm looking to create an app that uses an image uploaded by a user and looks for same/similar images on the web. Google images is the best example of that so I tried using the Vision API to first read the image labels and query them using Google Custom Search API.
The issue is, the label reading API doesn't give specific labels. For example, If I upload a starbucks logo on google images, I would get labels "starbucks" and "logo". Using the same image to get labels from the API gives me all other irrelevant labels such as "font", "circle", "art" etc.
My question is, how come an API by the same company that uses a functionality on their website is giving me different results when I use it?
Related
So I want to make a bot that is similar to the twitter accounts like "2 images of... per day" without having to use images that I would have downloaded but using images on sites like DeviantArt and adding the source of the image in the tweet.
Thank you.
I am new to stripe api And updating created product from api
https://stripe.com/docs/api/products/update?lang=java
from the docs it is cleared that for images
A list of up to 8 URLs of images for this product, meant to be displayable to the customer. When I passing 3 images urls in array using java api, response was successful
But when I go back to stripe dashboard it is displaying only 1 image.
I just want to know that if one image is displaying then why list of images is there.
Stripe only uses the first image in the list for products like Checkout and the dashboard. The API accepts a list of images despite only using the first one for future Stripe products.
Our app implements the Google Photos API to allow our user to select the images from their Google Photos album and then edit them.
However, in some cases, the image retrieved from Google Photos API (with URL as BASE_URL=d) is corrupted, i.e., part of the image content is grayed out.
We can't replicate the issue on our phone, so it is difficult to tell whether the image data is already corrupted in the Google Photos album, or the data becomes corrupted during the internet transmission.
We've also implement the jpg data check (SOI = 0xffd8 and EOI = 0xffd9) for the download image data. For a jpg source that is originally corrupt (i.e., SOI = 0xffd8 but EOI != 0xffd9) and then upload to Google Photos, the corrupt criterion can be correctly triggered for the download data via Google Photos API. But for the jpg images that the contents are partially grayed out as mentioned in this issue, the criterion is never triggered.
Before using Google Photos API, we've used Picasa API for about 2 years. And the corrupt image issue never happens. Does anyone have the same issue with Google Photos API?
I'm having a similar issue through the Google Photos web site and contacted support about it. They said this is a known issue that started Mar 20th and that engineers are working on it.
More detail about what I ran into: I uploaded a photo via Backup & Sync (their desktop sync tool), which puts them in Google Drive, and have it set to replicate all pictures over to Google Photos. The photo looks fine in Drive, but the version in Google Photos is corrupted with a large block of the photo being a solid color.
I am currently able to use the instagram API to read the basic user profile. I am able to get the profile_picture from the returned result via the API call below.
https://api.instagram.com/v1/users/self/?access_token=ACCESS-TOKEN
However, I notice that the avatar image is extremely small - it is 150x150. And extremely pixelated if enlarged on a mobile device.
Is it possible for me to get the avatar image that is higher definition?
If so, what is the API that i need to use?
Anyone knows if it is possible to use the flickr api to search by a specific camera model?
The camera model is stored in the exif data, but I cpuld find a way to search after exif data.
Flickr attaches Tags based on EXIF info in images. One of such tags can help you search based on camera.
Flickr's serach service (http://www.flickr.com/services/api/flickr.photos.search.html) has an option of search by tags.
This is an example of such searches: www.flickr.com/cameras/