Week #1 #

Team Formation and Project Proposal #

Team Members #

Value Proposition #

Problem Statement #

Photographers have a multitude of responsibilities, from capturing the perfect moment to selecting the best shots (a process called culling) and editing them. Culling can be an incredibly time-consuming task, as photographers may take thousands of shots in a day, especially at high-speed events like sports.

Solution Description #

Our team is developing a multiplatform tool called A-Shot that harnesses the power of computer vision and machine learning to expedite and simplify the culling process. A-Shot can detect blurry or overexposed shots and group similar shots together, all within an intuitive and user-friendly interface.

Benefits to Users #

A-Shot offers several benefits to photographers of all skill levels:

• Efficient Quality Checks: A-Shot’s blur and overexposure detection will quickly identify and eliminate low-quality shots, saving you time and effort.

• Simplified Comparisons: The tool’s similar shot grouping feature makes it easy to compare and select the best shot from a group of similar images.

• User-Friendly Interface: A-Shot’s intuitive interface simplifies the culling process and makes it less overwhelming, even when dealing with thousands of shots.

• Cost-Effective Solution: A-Shot is a budget-friendly solution for beginners and amateur photographers, as it does not require a subscription fee like other culling tools.

Differentiation #

A-Shot stands out from other culling tools in the market:

• Subscription-Free: Unlike tools like Capture One or Aftershoot, which can be expensive with subscription fees ranging from $10 to $100 per month, A-Shot is a cost-effective solution for all users.

• All-in-One Solution: Free viewers like FastRaw Viewer or FastStone Viewer lack the assistance algorithms that make culling faster and easier. While they can be used with grouping tools like PictureEcho, this isn’t as convenient as A-Shot’s all-in-one solution.

• Optimized for Culling: Cloud-based solutions like Google Photos may seem like an option, but they have limited free storage, and uploading thousands of pictures can be slow. Additionally, Google Photos might be not as convenient for culling as A-Shot.

User Impact: #

By using A-Shot, photographers can significantly reduce the time and effort spent on the culling process. This allows them to focus more on their creative work, such as capturing stunning moments and perfecting their editing skills.

Also, with A-Shot, photographers can deliver their best work to clients more quickly, improving their professional reputation and increasing the potential for repeat business and referrals.

Use Cases: #

• Use Case 1: A sports photographer uses A-Shot to quickly cull through thousands of action shots taken during a game. A-Shot’s blur and exposure imbalance detection helps the photographer eliminate low-quality shots, while the similar shot grouping feature allows for easy comparison and selection of the best images. The photographer can then quickly edit and submit the images to their client, meeting tight deadlines and showcasing their best work

• Use Case 2: An amateur photographer, passionate about capturing cityscapes and architecture, uses A-Shot to improve his/her photography skills and workflow. The cost-effective and user-friendly solution helps the photographer cull his/her images, learn from the mistakes, and identify the strengths. With A-Shot, the photographer can focus more on honing their craft and exploring new locations, rather than being bogged down by the time-consuming culling process.

Lean Questionnaire #

Please answer the following questions related to the lean methodology:

1. What problem or need does your software project address?

   A-Shot streamlines the photo culling process, saving photographers time and effort.

2. Who are your target users or customers?

   Target users are photographers of all skill levels, particularly those seeking a cost-effective solution.

3. How will you validate and test your assumptions about the project?

   We will validate and test our assumptions through user interviews, surveys, and an MVP. We will gather feedback and iterate the product based on the insights.

4. What metrics will you use to measure the success of your project?

   We will measure the success of A-Shot using user acquisition, growth, and satisfaction, and time saved in the culling process compared to other methods.

5. How do you plan to iterate and pivot if necessary based on user feedback?

   We will continuously gather and analyze user feedback, prioritize and implement changes in an agile and iterative manner, and be prepared to pivot our product strategy or business model if required to better address the needs of our target users.

Leveraging AI, Open-Source, and Experts #

Our team is prepared to leverage AI, open-source, and expert knowledge to develop A-Shot:

• Artificial Intelligence (AI): We will employ cutting-edge machine learning algorithms, such as convolutional neural networks (CNNs), and computer vision techniques to power A-Shot’s core features. Our team members, Timofey Brayko, Arthur Rakhmetov, Mikhail Romanov, and Artemii Miasoedov will focus on training and optimizing models for blur and exposure imbalance detection and similar shot grouping.

• Open-Source: We will take advantage of popular open-source libraries, frameworks, and tools to accelerate the development process and ensure the quality of our product. Specifically, we plan to use:

  • PyTorch: A machine learning framework for model training and deployment.
  • Jetpack Multiplatform: A toolkit for building multiplatform apps using Kotlin.
  • OpenCV: A computer vision library for image and video processing.

• Expert Knowledge: We will actively engage with photography experts and our target users to gather insights, validate our assumptions, and refine our product. Our UX/UI designer, Rusakov Matthew, will work closely with the users to ensure that A-Shot’s interface is intuitive, user-friendly, and tailored to their needs.

Defining the Vision for Your Project #

Overview #

A-Shot is a user-friendly, multiplatform tool that leverages computer vision and machine learning to simplify and accelerate the photo culling process for photographers. Within the given timeframe of several weeks, we will focus on the development of the following essential features:

1. Blur detection: To automatically identify and eliminate low-quality, blurry shots.

2. Exposure imbalance detection: To detect and remove images with poor exposure, ensuring that only the best-exposed shots are presented for selection.

3. Grouping of similar shots: To automatically group similar images together, making it easy to compare and choose the best shot from each group.

Schematic Drawings #

User flow demo #

Features matrix #

Tech Stack #

Our tech stack will include the following tools and technologies:

• PyTorch: For the training and deployment of machine learning models.
• Jetpack Multiplatform: For the creation of a multiplatform app using Kotlin.
• OpenCV: For image and video processing, as well as computer vision tasks.

Future Vision #

In the future, if time permits or as part of our post-launch development plan, we aim to incorporate additional features such as:

1. Face detection: To automatically identify and emphasize faces in images, aiding in the selection of shots with the best facial expressions.

2. Closed eyes detection: To assist photographers in quickly eliminating shots where subjects’ eyes are closed.

3. Image ranking by different properties: To automatically rank images based on factors like composition, exposure, and sharpness, further streamlining the culling process.

4. Grouping by faces: To automatically group images based on the identified faces, making it easier to select the best shots of each individual.

5. Any ideas that will appear during the development cycles: We will remain open to new ideas and innovations that emerge during the development process, and evaluate their potential for enhancing A-Shot’s capabilities and user experience.

These features will enhance A-Shot’s capabilities, making it an invaluable tool for photographers of all skill levels.