# CLVCNET

**Repository Path**: LeeWlving/CLVCNET

## Basic Information

- **Project Name**: CLVCNET
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Artistic-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2022-04-09
- **Last Updated**: 2022-04-10

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

This is the code for the paper:

"Comprehensive Linguistic-Visual Composition Network for Image Retrieval"

(1) FashionIQ:
Creat files "dress","shirt", and "toptee" to save the outputs.
dress: python train.py --dataset fashioniq --name dress --seed 599 --max_decay_epoch 20 --img_weight 1.0 --class_weight 1.0 --mul_kl_weight 1.0 --model_dir ./dress --num_epochs 50
shirt: python train.py --dataset fashioniq --name shirt --seed 599 --max_decay_epoch 20 --img_weight 1.0 --class_weight 1.0 --mul_kl_weight 1.0 --model_dir ./shirt --num_epochs 50
toptee: python train.py --dataset fashioniq --name toptee --seed 599 --max_decay_epoch 20 --img_weight 1.0 --class_weight 1.0 --mul_kl_weight 1.0 --model_dir ./toptee --num_epochs 50

(2) Shoes:
Creat files "shoes" to save the outputs.
python train.py --dataset shoes --seed 6195 --max_decay_epoch 30 --img_weight 1.0 --class_weight 1.0 --mul_kl_weight 1.0 --model_dir ./shoes --num_epochs 50

(3) Fashion200K:
Creat files "Fashion200k" to save the outputs.
python train.py --dataset fashion200k --seed 6195 --num 1 --img_weight 1.0 --class_weight 1.0 --mul_kl_weight 1.0 --model_dir ./fashion200k --num_epochs 40