Performance Report for August 2022
Name: Vedant Madane
Code: 1719
Branches worked on:
Branch Name | Details |
22-08-2022 | The branch that should be on staging |
30-08-2022 | Contains a three-way merge of flip vertically, face detection and signature validator. |
faceMatcher | FaceMatcher branch matches face in captured photo with face in photo with adhar |
loadModelsFromPublic | Loads our pre-trained machine learning models from the /public directory so as not to increase the bundle size of our /dist |
validSignature | Checks weather the uploaded images contains text, recognizes the text and says it is valid signature if it contains 5 to 10 characters. |
08-08-2022 | |
vite | Migrated from WebPack to vite |
uglyfy | Migrating from WebPack to uglyfy |
17-08-2022 | With model load & inference time console logs. |
mergeDetectFace | 09-08-2022 with all the commented code removed |
faceMatcher | Matches if person1 in photo is the same as person1 in photo with adhar. |
mediapipe | Need to select between cpu and WebGL Backends. |
09-08-2022 | Passing Functional Parameters, Face Detection Integration. Ready for Merging, Insourced Models. |
captureVideoFrame | Extract a frame from the captured video so that it can be passed to the model as prima facie the model is not accepting our video as input giving the above error: Uncaught (in promise) Error: toNetInput - expected media to be of type HTMLImageElement | HTMLVideoElement | HTMLCanvasElement | tf.Tensor3D, or to be an element id |
vladmandic | A fork of the FaceAPI package. |
20-07-2022 | Contains most up-to-date branch having Photo Capture in Profile Building |
blazeface | Model is loading but predicting randomly |
faceAPI | Model is not predicting as the code is in Vue3 |
XFaceDetector | XFaceDetector integration |
face-detection-vue | Vue3 setup() faceAPI integration into the uncomponentized Registration.vue page |
20-07-2022 | Photo Capture in Profile Building without Face Detection. |
Tesseract | Extract the adhar card number from from the webcam captured image |
219 hours were spent with a productivity pulse of 61%.
26% of this time was spent directly on software development whereas 15% was spent on learning on YouTube categorized here as Entertainment.
Productivity was 19% higher during work hours than during non-work hours.
Time spent on software development was 21% during work hours than during off-work hours.
About 28 hours were spent using Visual Studio Code for software development followed by 11 hours on localhost port number 8080. Github was used for prototyping and stackoverflow for doubt solving and error debugging. 111 hours were spent on all productive over 26 days out of the 31 days in August.
Less than 50 hours were spent on all distracting time which was lower than the 62 hours limit set.
33 hours were spent on software development cumulatively spread across VSCode and localhost. 4 hours was spent researching each on Github and Firefox whereas 6 and half hours were spent on communication and scheduling.
