Electronics Image Data Collection
Budget: $250 – $750 USD
I’m training a machine-learning model that must recognise consumer electronics, so I need a fresh, well-organised image dataset focused exclusively on product shots. The goal is to gather a wide visual range of smartphones, laptops, tablets, headphones, wearables, home-office gear and similar items photographed on plain backgrounds as well as in real-world settings.
Here’s the scope in a nutshell:
• Source or capture several thousand high-resolution JPEG or PNG images (at least 1024 px on the shortest side).
• Keep each file clearly named by product type and brand, then group the images in logical folders for easy ingestion into my training pipeline.
• Supply a simple CSV or JSON annotation file that pairs every filename with its product label; bounding boxes are a plus but not mandatory.
• Avoid watermarks, logos from unrelated brands, or duplicate angles that add no new visual information.
Once the full set is uploaded to a shared cloud drive of my choice and the annotation file passes a quick spot-check for accuracy, the project is complete.
Here’s the scope in a nutshell:
• Source or capture several thousand high-resolution JPEG or PNG images (at least 1024 px on the shortest side).
• Keep each file clearly named by product type and brand, then group the images in logical folders for easy ingestion into my training pipeline.
• Supply a simple CSV or JSON annotation file that pairs every filename with its product label; bounding boxes are a plus but not mandatory.
• Avoid watermarks, logos from unrelated brands, or duplicate angles that add no new visual information.
Once the full set is uploaded to a shared cloud drive of my choice and the annotation file passes a quick spot-check for accuracy, the project is complete.