Evolve AI Institute

Assessment Rubrics for Summative Evaluation

Lesson 7: How AI Sees Images

Lesson 7: How AI Sees Images

TABLE OF CONTENTS

  1. Image Recognition Process Diagram Rubric
  2. Written Explanation Rubric
  3. Comparison Chart Rubric (Human vs. Computer Vision)
  4. Real-World Application Analysis Rubric
  5. Teachable Machine Activity Rubric
  6. Class Participation Observation Rubric
  7. Optional Extension Project Rubric
  8. Vocabulary Understanding Rubric
  9. Holistic Understanding Rubric
  10. Self-Assessment Rubric for Students

RUBRIC 1: Image Recognition Process Diagram

Assignment: Create and label a diagram showing the four-step image recognition process

Total Points: 20

CriteriaExemplary (5 pts)Proficient (4 pts)Developing (3 pts)Beginning (2 pts)Not Yet (0-1 pt)
Accuracy of StepsAll four steps correctly identified and sequenced: Input → Feature Extraction → Pattern Matching → ClassificationAll four steps present with minor inaccuracies in details3 of 4 steps correct or sequence has errorsOnly 2 steps correct or significant conceptual errorsMissing steps or fundamentally incorrect process
Visual ClarityDiagram is exceptionally clear with logical flow, arrows, and visual organization; easy to followDiagram is clear and organized with adequate visual flowDiagram is somewhat organized but could be clearer; flow is present but confusingDiagram is disorganized; flow is unclear or absentNo clear diagram structure; incomprehensible layout
Labels and DescriptionsAll components labeled with detailed, accurate descriptions explaining what happens at each stepAll components labeled with adequate descriptionsMost components labeled but descriptions lack detail or have minor errorsFew labels; descriptions are vague or incorrectMissing labels or descriptions are mostly incorrect
Vocabulary UseCorrectly uses 5+ key terms: pixels, RGB, training data, features, patterns, confidence score, classificationCorrectly uses 4 key terms with appropriate contextUses 2-3 key terms, some may be used incorrectlyUses 1-2 key terms or uses them incorrectlyNo technical vocabulary or all terms used incorrectly

Comments:  

Total Score: ______ / 20

RUBRIC 2: Written Explanation - How AI Learns

Assignment: Write 1-2 paragraphs explaining how AI learns to recognize images using training data, with at least one specific example

Total Points: 20

CriteriaExemplary (5 pts)Proficient (4 pts)Developing (3 pts)Beginning (2 pts)Not Yet (0-1 pt)
Understanding of Training DataClearly explains that AI learns from thousands of labeled examples and why diversity matters; shows deep understandingExplains AI learns from examples; mentions importance of quantity/qualityMentions training data but explanation is superficial or partially incorrectVague reference to AI learning; doesn't clearly explain training dataNo understanding of training data concept evident
Specific ExampleProvides detailed, relevant example with clear explanation of how training works (e.g., "showing AI 10,000 dog photos of different breeds...")Provides adequate example that illustrates conceptProvides example but it's generic or doesn't clearly illustrate conceptAttempts example but it's incorrect or irrelevantNo example provided
Writing QualityWell-organized paragraphs with topic sentences, clear transitions, and strong conclusion; 1-2 paragraphs as requestedOrganized writing with clear main ideas and adequate developmentBasic organization; ideas are present but development is weakPoorly organized; ideas are disconnected or unclearNo clear organization; writing is incomprehensible
Technical AccuracyAll statements are scientifically accurate; no misconceptionsMostly accurate with only minor errorsSome accurate information but contains misconceptionsSignificant misconceptions or inaccuraciesFundamentally incorrect understanding

Comments:  

Total Score: ______ / 20

RUBRIC 3: Comparison Chart - Human vs. Computer Vision

Assignment: Create a comparison showing at least three differences between human vision and computer vision

Total Points: 15

CriteriaExemplary (5 pts)Proficient (4 pts)Developing (3 pts)Beginning (2 pts)Not Yet (0-1 pt)
Number of DifferencesIdentifies 5+ meaningful differences with specific detailsIdentifies 4 meaningful differencesIdentifies 3 differencesIdentifies only 1-2 differencesIdentifies no clear differences or all are incorrect
Depth of AnalysisEach difference includes explanation of why/how with specific examples; shows sophisticated understandingEach difference explained clearly with some detail or examplesDifferences listed but explanations are basic or missing detailDifferences listed with little to no explanationNo meaningful analysis; just words without understanding
AccuracyAll comparisons are accurate and demonstrate clear understanding of both systemsMostly accurate; minor errors don't affect overall understandingSome inaccuracies but core understanding is presentSeveral significant errors or misconceptionsMostly or entirely inaccurate

Comments:  

Total Score: ______ / 15

RUBRIC 4: Real-World Application Analysis

Assignment: Identify one use of image recognition in your life and explain one benefit and one concern

Total Points: 15

CriteriaExemplary (5 pts)Proficient (4 pts)Developing (3 pts)Beginning (2 pts)Not Yet (0-1 pt)
Application IdentificationIdentifies specific, relevant real-world application with details about how it worksIdentifies clear real-world application with adequate descriptionIdentifies application but description is vague or genericApplication mentioned is questionable or very genericNo clear application identified or entirely irrelevant
Benefit ExplainedBenefit is clearly explained with specific reasons why it's helpful; shows understanding of impactBenefit identified and explained with adequate reasoningBenefit mentioned but explanation is superficialBenefit stated but not explained or explanation doesn't make senseNo benefit identified or completely incorrect
Concern ExplainedConcern is thoughtful and specific; may include privacy, bias, accuracy, or ethical considerations; well-reasonedValid concern identified with reasonable explanationConcern mentioned but explanation is basic or unclearConcern is vague or seems unrelated to the technologyNo concern identified or completely irrelevant

Comments:  

Total Score: ______ / 15

RUBRIC 5: Teachable Machine Activity Performance

Assignment: Successfully train an image classification model and document process/results

Total Points: 20

CriteriaExemplary (5 pts)Proficient (4 pts)Developing (3 pts)Beginning (2 pts)Not Yet (0-1 pt)
Model TrainingSuccessfully trained model with 30+ diverse images per class; model performs wellTrained model with adequate images (20-30); model works reasonably wellTrained model but limited images (<20) or limited diversity; model has inconsistent performanceAttempted training but model doesn't work well due to insufficient or poor-quality imagesDid not successfully train a model
Testing and ObservationExtensively tested model with various scenarios; documented what works well and what causes errorsTested model adequately; documented basic observationsLimited testing; minimal documentationVery limited testing; little to no documentationNo testing or documentation
Reflection QualityInsightful reflections on why AI succeeded/failed; connects to lesson concepts; identifies patternsGood reflections with clear observations about AI behaviorBasic reflections; some observations but lack depthMinimal reflection; observations are superficialNo meaningful reflection
Worksheet CompletionWorksheet fully completed with thoughtful, detailed responsesWorksheet completed with adequate responsesWorksheet partially completed or responses lack detailWorksheet minimally completed; most responses incompleteWorksheet not completed or responses are blank/invalid

Comments:  

Total Score: ______ / 20

RUBRIC 6: Class Participation Observation

Use throughout lesson; teacher observation notes

Total Points: 10

CriteriaExemplary (3-4 pts)Proficient (2 pts)Developing (1 pt)Not Yet (0 pts)
EngagementActively engaged throughout; asks questions; makes connections; shows curiosityGenerally engaged; participates when asked; follows alongSometimes engaged; frequently off-task or passiveRarely engaged; consistently off-task
CollaborationWorks exceptionally well with partner/group; contributes ideas; listens to others; helps peersWorks well with others; shares tasks fairly; communicates adequatelyStruggles with collaboration; may dominate or disengage from groupDoes not collaborate effectively; conflicts or complete disengagement
Use of Class TimeUses time efficiently; stays on task; completes activities; helps others when finished earlyUses time adequately; completes most activities with appropriate effortUses time poorly; off-task frequently; rushes through activitiesWastes time; does not complete activities; distracts others

Observation Notes:  

Total Score: ______ / 10

RUBRIC 7: Optional Extension Project

Assignment: Create video explanation, AI system proposal, or other extension project

Total Points: 30 (Extra Credit or Alternative Summative)

CriteriaExemplary (6 pts)Proficient (5 pts)Developing (3-4 pts)Beginning (1-2 pts)Not Yet (0 pts)
Content AccuracyAll information is accurate, detailed, and demonstrates deep understandingContent is mostly accurate with good understanding demonstratedContent has some inaccuracies or shows partial understandingContent has significant errors or misconceptionsContent is mostly incorrect or incomplete
Creativity/OriginalityHighly creative approach; original thinking; goes beyond lesson contentShows creativity; fresh ideas or perspectivesSome creative elements but mostly follows standard approachesLittle creativity; generic or copied approachNo creativity; minimal effort evident
OrganizationExceptionally well-organized and easy to follow; logical flow; professional qualityWell-organized with clear structure and adequate flowBasic organization; structure is present but could be improvedPoorly organized; hard to follow; lacks structureNo clear organization; incomprehensible
Technical ExecutionHigh-quality production; appropriate use of tools/media; polished final productGood quality; adequate use of tools; complete final productAdequate quality but technical issues or incomplete elementsPoor quality; significant technical problemsVery low quality or unfinished
Depth of ThinkingShows sophisticated analysis, synthesis, or problem-solving; addresses complex questionsShows good thinking skills; makes connections; adequate depthShows basic thinking; surface-level analysisShows limited thinking; very superficialShows no meaningful thinking or analysis

Comments:  

Total Score: ______ / 30

RUBRIC 8: Vocabulary Understanding

Assignment: Demonstrate understanding of key terms through matching, definition writing, or usage in context

Total Points: 16 (2 points per term)

Terms to Assess: Pixel, Pattern Recognition, Training Data, Classification, Computer Vision, Feature Extraction, Confidence Score, Machine Learning

TermFully Understands (2 pts)Partially Understands (1 pt)Does Not Understand (0 pts)
PixelCorrectly defines as tiny colored square with RGB values; explains role in digital imagesBasic definition but missing key detailsIncorrect or no definition
Pattern RecognitionExplains as process of finding recurring features/arrangements in data; connects to AIBasic definition but incomplete or vagueIncorrect or no definition
Training DataExplains as set of labeled examples used to teach AI; mentions importance of diversityBasic definition but missing key conceptsIncorrect or no definition
ClassificationExplains as assigning category/label based on features; understands it's AI's outputBasic definition but incompleteIncorrect or no definition
Computer VisionExplains as AI field enabling computers to interpret visual informationBasic definition but vagueIncorrect or no definition
Feature ExtractionExplains as identifying important characteristics (edges, colors, shapes)Basic definition but incompleteIncorrect or no definition
Confidence ScoreExplains as percentage showing AI's certainty about predictionBasic definition but incompleteIncorrect or no definition
Machine LearningExplains as AI learning patterns from examples vs. following rigid rulesBasic definition but incompleteIncorrect or no definition

Comments:  

Total Score: ______ / 16

RUBRIC 9: Holistic Understanding - End of Unit

Use for overall assessment of mastery

Total Points: 25

LevelDescriptionPoints
Advanced (22-25 pts)Student demonstrates comprehensive understanding of how AI processes images, including technical details and real-world implications. Can explain concepts clearly to others, connects ideas across lessons, and thinks critically about ethical implications. Shows ability to apply knowledge to new situations. Consistently uses technical vocabulary correctly.22-25
Proficient (18-21 pts)Student demonstrates solid understanding of core concepts. Can explain the image recognition process, role of training data, and differences between human and computer vision. Understands real-world applications and basic ethical concerns. Uses vocabulary correctly most of the time. May need support with more complex applications or connections.18-21
Developing (14-17 pts)Student demonstrates partial understanding of concepts. Grasps basic ideas about pixels and how AI processes images but may have gaps in understanding training data importance or real-world applications. Uses some vocabulary correctly but inconsistently. Needs support to explain concepts clearly or make connections.14-17
Beginning (10-13 pts)Student demonstrates limited understanding. Recognizes that AI processes images differently than humans but cannot explain details accurately. Struggles with vocabulary and has difficulty making connections. May confuse key concepts. Needs significant support and reteaching.10-13
Not Yet (0-9 pts)Student shows minimal or no understanding of core concepts. Cannot explain how AI processes images or why training data matters. Does not use vocabulary correctly. Needs intensive intervention and may require alternative instruction approaches.0-9

Supporting Evidence:  

Total Score: ______ / 25

RUBRIC 10: Student Self-Assessment

For students to evaluate their own learning

Name:   Date:  

StatementStrongly Agree (4)Agree (3)Disagree (2)Strongly Disagree (1)
I understand how AI processes images as pixel data
I can explain what training data is and why it matters
I can describe the four steps in image recognition
I understand differences between human and computer vision
I can explain at least one real-world use of image recognition
I understand ethical concerns about facial recognition
I can use vocabulary like pixel, classification, and training data correctly
I worked well with my partner/group during activities
I asked questions when I was confused
I am interested in learning more about computer vision

Total Points: ______ / 40

What was easiest for you in this lesson?

What was most challenging?

What would help you understand better?

How can you use what you learned?

SCORING GUIDE FOR TEACHERS

Overall Lesson Assessment Score

If using all rubrics, calculate total possible points and student score:

Total Possible: 20 (Diagram) + 20 (Written) + 15 (Comparison) + 15 (Application) + 20 (Teachable Machine) + 10 (Participation) + 16 (Vocabulary) = 116 points

Grading Scale:

Flexible Assessment Options

Option 1: Select 3-4 Rubrics

Choose the rubrics most aligned with your learning objectives and grade only those.

Option 2: Weight Rubrics Differently

Assign different weights based on importance (e.g., Diagram = 30%, Teachable Machine = 30%, Written = 20%, Application = 20%)

Option 3: Portfolio Assessment

Collect all work in a portfolio; assess overall growth and understanding holistically using Rubric 9.

Option 4: Mastery-Based

Students must achieve "Proficient" or higher on core rubrics (Diagram, Written, Teachable Machine) to demonstrate mastery. Reteach and reassess if needed.

FEEDBACK TEMPLATES

For Exemplary Work:

"Excellent work! Your [specific element] demonstrates deep understanding of [concept]. I especially appreciate [specific strength]. To challenge yourself further, consider [extension idea]."

For Proficient Work:

"Good job! You clearly understand [concept]. Your [specific element] shows solid thinking. To strengthen your work, try [specific suggestion]."

For Developing Work:

"You're making progress on understanding [concept]. I can see you understand [strength], but [specific area] needs more work. Let's review [concept] together. Try [specific strategy]."

For Beginning Work:

"I can see you're working on understanding [concept]. Let's work together to strengthen [specific area]. Here's what will help: [specific intervention]. Come see me so we can review [concept]."

DIFFERENTIATED ASSESSMENT OPTIONS

For Advanced Students:

For Struggling Students:

For ELL Students:

For Students with IEPs:

USING RUBRICS EFFECTIVELY

Before Lesson:

During Lesson:

During Assessment:

After Assessment:

Evolve AI Institute • Lesson 7: How AI Sees Images

Assessment Rubrics for Summative Evaluation

Remember: Rubrics should support learning, not just measure it. Use them as tools for clear communication of expectations, specific feedback, and growth tracking. Adapt them to your students' needs and your teaching context!